Euromedim 2006 :1st European Conference on Molecular Imaging Technology

9 May 2006, 08:00 → 12 May 2006, 18:00 Europe/Zurich

Palais du Pharo, Marseille

The main aim of this conference is to develop cross-fertilization between different disciplines and to attract in a same place medical doctors, medical physicists and physics detector experts as well as industrial partners to discuss about needs and improvements to be made in the area of clinical nuclear medical imaging like Single Photon Computed Tomography (SPECT) and Positron Emission Tomography (PET).

Tuesday 9 May

08:30 → 09:30 Conference Registration

09:30 → 10:30 Official ceremony

chaired by Professor Y. Berland, President of "Université de la méditerranée" under the honorary presidence of the Professor J. F. Mattei, former french Minister of health, president of French Red-Cross

11:00 → 12:30 Opening session

11:00 Impact of fundamental Science on Society

Speaker: Dr Robert Aymar (CERN Director General)

Slides Tue1100-P1.pdf Tue1100-P1.ppt

11:30 Impact of Imaging on life Science

Speaker: Dr Andre Syrota (CEA/DSV)

12:00 A Brief History of LSO: From Conception to Commercial Implementation

The evolution of Lu2SiO5:Ce from an experimental powder phosphor synthesized in the laboratory to the large scale commercial production of single crystals implemented in Positron Emission Tomography spanned more than 10 years and required solutions to numerous technical problems including the understanding of the scintillation mechanism, the purification of raw materials, the development of a practical crystal growth process, the development of detector processing techniques, the optimization of detector design, and the integration of detectors into imaging systems. It also required financial commitment, patience, luck, and a willingness to accept risk. The discovery and development of LSO:Ce are discussed from both historical and technical points of view, and prospects for the use of LSO:Ce in future advanced imaging systems such as Time-of-Flight PET are considered in comparison to new scintillators that are on the horizon.

Speaker: Prof. Chuck Melcher (The University of Tennessee)

14:00 → 15:00 Poster Session :Simulation, Modeling, Reconstruction

14:00 3-D Reconstruction from Compton Scattered Data Using Two Different Approaches to Implementing Projector-Backprojector Pair

Iterative image reconstruction for Compton camera is computationally challenging since the projection and backprojection are performed on conical surfaces rather than along straight lines and there is a number of possible combinations of position and energy measurements. Therefore, computationally efficient implementation of a projector-backprojector pair without loss of quantitative accuracy is required. Methods: Two different approaches to conical surface integral were investigated for rapid calculations of an approximately factorized system matrix used for 3-D EM (expectation maximization) reconstruction: the ellipse-stacking method (ESM) and the ray-tracing method (RTM). For every possible position in two detectors and scattering angles, the axis and half angle of the conic surface were determined. Elements for system matrix were then calculated by the product of differential cross-section of Compton scattering in scatterer and the probability with which the voxel belongs to a given conic surface. The belonging probability for ESM was determined by the distance from the voxel center to the ellipse on a plane parallel to the scatterer. For RTM, the belonging probability was determined by the intersection length of the voxel with a straight line through the apex of the cone. It was assumed that Compton camera consisted of a pair of parallel scatterer and absorber with 16x16 detector elements in 5x5 cm2 active areas spaced by 5 cm. The scattering angle of the incident photon at the scatterer was quantized into 30 discrete angles between 10 and 100 degrees. Projection data for point source and three-cylinder phantom were simulated using both ESM and RTM. The iterative EM reconstruction algorithms were also implemented using both ESM and RTM. Computation time (for projections) and percent error between the mathematical phantom and reconstructed image (64x64x64 matrix with pixel size of 1.56 mm) were compared. Results: Computation time for projection using ESM was approximately 3 times longer than RTM. Percent errors (normalized root-mean squared errors) measured at the 30th iteration for three-cylinder phantom were 44% for ESM and 39% for RTM. Point source images were well reconstructed by both methods: FWHMs measured by the profile across the point image reconstructed with 30 iterations were 1.0 pixels for both methods in the plane parallel to the scatterer, and were 1.3 and 1.8 pixels for ESM and RTM, respectively, in the plane perpendicular to the scatterer. Conclusions: Two different approaches to implementing projector-backprojector pairs for 3-D reconstruction from Compton scattered data were developed. While RTM was more efficient than ESM in computation time, accuracies of the reconstructions by both methods were equivalent. The use of geometrical symmetry along with an efficient caching scheme could greatly reduce the computation time for both methods.

Speaker: Ms Soo Mee Kim (Department of Nuclear Medicine, Seoul National University College of Medicine)

14:00 A NEW APPROACH TO CALCULATION OF PLANAR DETECTOR EFFICIENCY FACTORS FROM COINCIDENCE DATA FOR THE NON-ROTATING QUAD-HIDAC PET SCANNER

The very high resolution quad-HIDAC PET scanner has four detector banks each with four planar high density avalanche chambers, forming 16 modules in total. The scanner has a large field of view (FOV) of 170 mm x 170 mm x 280 mm able to do whole body rodent imaging. It has many important applications, just one example being dynamic neuroreceptor imaging in mice. However, detector normalisation is crucial, meaning that degrading factors introduced by the scanner’s geometry and detector efficiencies must be corrected. To see the effects on a reconstructed image, a list-mode data set from a 18F uniformly filled cylindrical phantom (3cm in diameter and 5cm in length) was reconstructed, which revealed problems in the central transverse slices and a square shape due to the gaps between the planar detectors. In order to estimate the detector efficiencies a specific method for the quad-HIDAC scanner is proposed. The approach creates four matrices, each representing a bank of panel converters, such that each matrix element represents several stacked detector elements. The method fits a 2nd order polynomial to each row of each of these matrices in order to reduce noise, thus obtaining a better efficiency distribution for each detector bank. In order to validate the method a Monte Carlo simulation was developed. Different detector efficiency distribution models were tested and events from a point source using these detection efficiencies were simulated and recorded within a list-mode data file. The objective is recover the efficiencies using only the coincidence events collected. The simulation shows that the proposed method gives a good approximation for each model tested. The efficiencies ei-ej, joining each line of response (LOR), were included within an expectation maximisation (EM) reconstruction algorithm, which also includes a scanner geometry normalisation. Various reconstruction approaches were tested to see the difference between corrected images and uncorrected ones. The results show a notably improved homogeneity in the reconstructed images across the transaxial, coronal and sagittal views and an appreciable correction for the gaps between detectors banks. The reconstructed images were corrected without using scatter or attenuation corrections, which are relatively minor for this size of phantom.

Speaker: Ms Leticia Ortega Maynez (School of Chemical Engineering and Analytical Science at the University of Manchester)

14:00 A New Reconstruction Method for 3-D PET named pair of near-missing LORs ( PNMLOR) method

Traditionally, three dimensional (3-D) positron emission tomography (PET) image reconstruction is performed with the filtered back projection (FBP) method or the iterative reconstruction method [1,2]. These methods do not directly apply the geometrical property of lines of response (LORs) whereby two LORs, which originate from radioactive isotopes in the same position, lie within a few millimeters distance of each other. We proposed, PNMLOR, a new image reconstruction method of 3- D PET. The proposed method makes direct use of the elementary property of LORs, which originate from radioactive isotopes in the same position. As an application of the PNMLOR method, we present image reconstructions of a simulation of a portable PET. GATE was used for simulation. In the simulation, CdTe was assumed for the gamma ray detector. Each CdTe detector element was assumed to have a cross-section of 2 mm x 2 mm and a depth of 5 mm. The detector array was arranged as a cylinder having a diameter of 22 cm and a length of 20 cm.The CdTe detector elements were placed all around the lateral face of the cylinder. Digital Hoffmann phantom was used for a simulation. A simulation using the proposed method demonstrated good resolution.

Speaker: Dr Shoji Kawatsu (Kyoritu General Hospital, National Center for Geriatrics and Gerontology)

14:00 Analytical Reconstruction of Dynamic Cardiac SPECT with Noise-Reduction and Attenuation Compensation

Dynamic or gated cardiac SPECT imaging has been playing a major role in evaluation of chest pain and other cardiac symptoms. Conventionally, the acquired dynamic data sequences are reconstructed frame by frame, where the temporal correlation among different frames is not utilized. In this work, we proposed a framework for analytical reconstruction in the Karhunen-Loeve (KL) domain of quantitative dynamic SPECT with both noise reduction and accurate compensation for non-uniform attenuation. The dynamic data sequences were first transformed into the KL domain, in which the spatio-temporal reconstruction of the entire sequences becomes a series of reconstruction on the independent KL components. To suppress noise propagation through the reconstruction, a penalized weighted least-squares (PWLS) filtering was applied and then the filtered KL components were reconstructed one-by-one based on the Novikov’s inverse formula which compensates for non-uniform attenuation. By the proposed framework, a filtered back-projection (FBP)-type inversion for quantitative dynamic SPECT is feasible with accurate compensation for non-uniform attenuation and effective treatment of Poisson noise. Our computer simulation results are very encouraging, by both qualitative and quantitative judgments, as compared to both the frame-by-frame Novikov inversion and the conventional FBP reconstruction.

Speaker: Prof. Hongbing Lu (Department of Computer Application, Fourth Military Medical University, Shaanxi, China)

14:00 Assessment of input function distortions on kinetic model parameters in simulated dynamic 82Rb PET perfusion studies

Objectives: The determination of the exact shape of the input function is a particular challenge in dynamic cardiac 82Rb PET perfusion studies. The dynamic range of the activity in these studies easily covers two orders of magnitude. Additionally, the injection bolus has a very sharp peak. The measurement of the input function depends on both experimental errors (e.g. random noise, limited spatial resolution) and data processing (e.g. frame durations). These issues are a source of error in the assessment of the input function, especially regarding the peak value and the decaying slope. Thus, the goal of our study is to address the influence of such distortions of the input function on the model parameter K1, which is linked to myocardial perfusion [1]. Methods: Using a one-compartment kinetic model [2,3,4], noise-free myocardium TACs representing dynamic PET datasets were simulated with model parameters covering a range of physiologic interest [3]. The TAC simulations are based on different input functions computed with a generic analytical function. The parameters of this function have been derived by numerical fits of various Rubidium PET measurements of the left ventricular bloodpool. The simulated input function and myocardium TACs have then been applied in a kinetic analysis in the following way: the analytical input function has been distorted by parameter variation, leading to different peak values and decaying slopes. Then, the kinetic parameters have been re-estimated, using the respective (unmodified) myocardium TAC as reference data. The (relative) bias of the K1 estimation was calculated and analyzed as a function of the input TAC variation, and compared across the investigated range of kinetic model parameters. Results: Underestimating (overestimating) the input peak value causes an overestimation (underestimation) of K1, respectively. The magnitude of this effect depends strongly on the blood volume fraction, and the FWHM of the input function. This is because the wider the input peak, the stronger is the coupling of the input function to the observed myocardium TAC. For reasonable values of the model parameters, the relative bias in K1 is easily ±(10-30)% for a ±10% error in the input peak. This causes an even larger bias in the blood flow values, due to its nonlinear coupling to K1 [5]. Conclusion: Even with noise-free data, moderate errors in the estimation of the input peak value lead to significant errors in the estimated K1 parameter. Therefore, an accurate estimation of the input peak, e.g. by appropriate frame durations, is necessary for a reliable kinetic analysis and blood flow estimation. References: [1] M. E. Phelps, “PET – Molecular Imaging and Its Biological Applications”, Springer-Verlag, New York, 2004 (chapter 6) [2] P.G. Coxson, R.H. Huesman, L. Borland, J. Nucl. Med. 1997; 38:660-667 [3] L. Golanowski, R. A. de Kemp, R. S. Beanlands, T. D. Ruddy, Proc. 22nd EMBS Intern. Conf., July 23-28, Chicago IL, pp. 1096-1099 [4] G. El Fakhri et al, J. Nucl. Med. 2005; 46:1264-1271 [5] E. M. Renkin, Am. J. Physiol. 1959;197:1205-1210, and C. Crone, Acta Physiol. Scand. 1964;58:292-305.

Speaker: Mr Andreas Thon (Philips Research Laboratories)

14:00 AUTOMATIC CALCULATION OF SUV ON FDG PET-CT LUNG LESIONS

Aim: A new method for the automatic calculation of the mean standardized uptake value (meanSUV) in pulmonary lesions detected by FDG PET-CT has been developed and validated. Materials and methods: In PET-CT, calculation of FDG meanSUV rests on the delineation of a region of interest (ROI) around the tumor on a PET section. We propose to draw the ROI by utilizing the limits of the lesion as seen on the matched CT section. Once the tumor has been selected, a segmentation based on a fixed density threshold automatically delineates a ROI around the tumor on the CT section. This ROI is then applied to the PET image to calculate the meanSUV. This method was applied to a section passing through the center of a pulmonary masse or solitary nodule obtained in a series of 56 patients imaged with a PET/CT (Discovery ST,GE Health Care) one hour after injection of 340-440 MBq of FDG. The automatic method was compared with the manual method in which the operator manually selected the best-suited circular ROI to surround the tumour on the PET image. ROI size was also measured. All measurements were performed independently by two operators. Results: With the manual method, meanSUV and ROI size were respectively (m±sd) 3.47±1.58 and 73±54 for the first operator vs 3.83±1.61 and 64±53 (p<0.001 for both) for the second one. Between the two operators, the coefficient of correlation was 0.95 for the meanSUV and 0.96 for the ROI size. With the automatic method, the meanSUV and ROIsize found by the two operators were always the same for all the patients : they were respectively 4.40±2.22 and 65±50. In both cases, the coefficient of correlation was one. The comparison of the two methods by a same operator gave the next results : meanSUV and ROIsize were respectively 3.85±1.67 and 63±52 for the manual method versus 4.40±2.22 and 65±50 for the automatic method (p<0.001 for the meanSUV). The coefficient of correlation was 0.92 for the meanSUV and 0.93 for the ROIsize.The higher meanSUV obtained with the automatic, CT-based method can be explained by a better delineation of the tumor when compared with a circular ROI Conclusion: The use of an automatic, CT-based ROI to surround a tumour on a FDG PET image provides highly reproducible SUV values.

Speaker: Mr Ismael KARIDIOULA ((1) Centre Jean Perrin, Division of Nuclear Medicine, F-63011 Clermont-Ferrand, France, (2)University of Auvergne, Faculty of Medicine, F-63000 Clermont-Ferrand, France, (3)INSERM, U484, F-63000Clermont-Ferrand, France)

14:00 Comparison of simulated and experimental performances of the Concorde MicroPET Focus 220 system using the SORTEO Monte Carlo simulation software

Objectives: Recent advances in PET systems dedicated to small animal imaging have provided new ways to perform biologic research. They consecutively increased the need to develop realistic simulation models that allows optimizing acquisition and reconstruction protocols in order to improve image quantification and detection. PET-SORTEO is a Monte Carlo-based simulator that enables the fast generation of realistic PET data for the geometry of cylindrical PET scanners. It has been shown to accurately reproduce data of the human ECAT EXACT HR+ scanner and of the rodent CTI-Concorde MicroPET R4 system. Our aim is to adapt and configure this simulation tool for the MicroPET Focus 220, which belongs to the last generation of rodent and primate systems. Our originality, as compared with a recent simulation study using the Gate platform based on the Geant4 toolkit package, is that we demonstrate the feasibility to achieve realistic simulations of whole-body biological distributions in a very short computation time. Method: This validation study is carried out against actual measurements either performed on the actual Focus 220 scanner of the ARC Seibersdorf center or using performances measurements available in the literature, following the experimental protocol by Tai et al. The comparison of simulated and experimental performance measurements includes spatial resolution, sensitivity, energy spectra, scatter fraction and count rates. Realistic simulated whole-body mice acquisitions are also generated using the Moby phantom from Segars et al with different acquisition configurations, including dynamic studies. Realistic time activity curves are derived from experimental acquisitions on the Focus system of the ARC Seibersdorf center. Results: Preliminary results were achieved for the comparison of spatial resolution measurements. Experimental data were acquired using a 18F point source embedded in a 0.3 mm inner diameter capillary tube. The point source was located at the centre of one detector block and at the centre of the transverse FOV and moved in the vertical direction by steps up to the edge of the FOV (100 mm). Acquisition at each step was performed with an energy window of 250–750 keV and a timing window of 10 ns. Simulated data were Fourier rebinned (FORE) to form 2D sinograms and subsequently reconstructed using 2D FBP with a ramp filter cutoff at the Nyquist frequency and OSEM with 4 iterations and 16 subsets. The image spatial resolution was measured as the FWHM and the FWTM of a Gaussian profile fit to the measured resolution profile. These experimental data were compared to simulated data that were generated following a similar acquisition and reconstruction protocol. Experimental and simulated data are well superimposed with a discrepancy lower than 5%. Experimental data were also acquired to measure the system sensitivity, scatter fraction and NEC rates. Our ongoing work is to simulate equivalent acquisition protocols. Comparative results will be presented at the conference together with realistic simulated mice acquisitions.

Speaker: Dr Carole Lartizien (Creatis laboratory, UMR CNRS 5515 et INSERM U630, Villeurbanne, France)

14:00 CONTRAST NOISE BEHAVIOUR FOR A ROTATING SLAT COLLIMATED GAMMA CAMERA

INTRODUCTION Traditional gamma camera imaging with a parallel hole collimated detector (PH) is limited by the sensitivity versus spatial resolution tradeoff which is intrinsic to the design. This results in a reduced contrast for small lesions. The higher geometrical sensitivity of a rotating slat (RS) collimated strip detector could improve the image quality considerably. By means of phantom measurements, this study investigates if a RS gamma camera, with a spatial resolution of 5 mm at 10 cm collimator distance, is able to gain an improved contrast/noise ratio compared to a traditional camera, equipped with a LEHR collimator. METHODS Projections measured with a RS collimated gamma camera are planar integrals of the activity distribution and contain less information compared to ray projections, acquired with a traditional parallel hole collimated gamma camera. Therefore the RS collimated detecor has to spin around its own axis in order to collect complete data. Since planar projections are collected at different spin angles, image reconstruction comparable to a regular SPECT reconstruction is needed to obtain planar images. In this study we use a Monte Carlo based model of the acquisition physics. Incorporation of this model in the forward and backward projection of the MLEM algorithm yielded a Monte Carlo based reconstruction technique (MLEM-MC). A 17 cm diameter disc of uniform activity was printed on a sheet of paper together with 12 hot lesions located inside the uniform disc. The lesions had diameters ranging from 4 to 20mm and a contrast of 4:1. The sheet was placed at 10 cm from the collimator and 10 different (400 seconds) acquisitions were performed on both the PH and the RS camera. The data from the RS device were reconstructed in 250 iterations using MLEM-MC. The images coming from the PH camera were interpolated to have the same pixel size as the reconstructed RS images (1,8x1,8mm). The contrast recovery coefficient (CRC), defined as the ratio of the lesion activity and the background activity, was calculated for each lesion over the 10 different realisations of the measurement. Afterwards, the mean CRC (mCRC) over all lesions was calculated. For each pixel, the pixel noise was calculated as the standard deviation divided by the mean value over the 10 realisations. Averaging over all pixels yielded a global noise level (NL) of a measurement. RESULTS With the NL matched, the RS image reconstructed with MLEM-MC (60 iterations) reached a 21,0% higher CRC for the smallest hot spot, a 7,6% higher CRC for the 20mm hot lesion and a mCRC that was 20,0% higher compared to the PH values. For an equal mCRC, the NL of the RS image was 40,3% lower compared to the PH NL. For obtaining the same NL on a parallel hole system, the measurement takes 2,81 times longer. CONCLUSIONS When the same imaging time is used for all acquisitions, a lower NL is obtained for RS images at the same mCRC. On the other hand, a higher mCRC can be obtained when looking at images of the same NL. For comparable image quality on both modalities, the imaging time could be significantly reduced using the RS system.

Speaker: Mr Roel Van Holen (Ghent University ELIS-MEDISIP)

14:00 Detector normalization and scatter correction for the jPET-D4: a 4-layer depth-of-interaction PET scanner

The jPET-D4 is a brain PET scanner composed of 4-layer depth-of-interaction (DOI) detectors with a large number of GSO crystals, which achieves both high spatial resolution and high scanner sensitivity. Since the sensitivity of each crystal elements highly depends on the depth of DOI layers and the energy of an incident gamma ray, it is difficult to estimate normalization factors and scatter components with high statistical accuracy. In this work, we implemented a hybrid scatter correction method combined with component-based normalization, which estimates scatter components with a dual energy acquisition through an estimation of trues from an upper energy window using a convolution subtraction method. In order to reduce statistical noise in sinograms, the implemented scheme uses the DOI compression (DOIC) method, which combines deep pairs of DOI layers into the nearest shallow pairs of DOI layers with natural detector samplings. Since the compressed data preserve block detector configuration as if acquired with ‘virtual’ detectors having a high stopping power for the gamma rays, the correction methods can be applied directly to DOIC sinograms. To investigate the performance, we compare the DOIC based correction with the conventional non-DOI correction using phantom measurements of the jPET-D4. The results indicate that the quality of normalization and scatter correction of the DOIC data is almost the same as those of the non-DOI data, while the statistical error in uncompressed DOI data increases rapidly with the depth of DOI layers. The proposed method also allows us to improve computational efficiency and to reduce acquisition time and storage size for normalization data even for multi-layer DOI detectors.

Speaker: Dr Keishi Kitamura (Shimadzu Corporation)

14:00 Effect of Detector Parameters on the Image Quality of Compton Camera for 99mTc

The Compton camera, which uses a new concept of collimation, i.e., electronic collimation, has a bright future as a medical imaging device considering its simplicity and inherent 3D image capability. Currently, however, the spatial resolution of the Compton camera is not sufficient for medical imaging. In this study, we investigated the effect of various parameters in Compton camera on image quality by using a general purpose Monte Carlo simulation package GEANT4. This study focused on the low energy gamma source, 99mTc, which emits 140 keV gammas. This study modeled a Compton camera which consists of two plane-type position-sensitive detectors: a double-sided silicon strip detector (DSSD, 5 x 5 x 0.15 cm3) as scatterer and a 25-segmented germanium detector (25-SEGD, 5 x 5 x 2 cm3) as absorber. The distance between the scatterer and absorber is 5 cm. This study modeled a 99mTc point source at 6 cm from the scatterer. The Compton camera was modeled very realistically in this study including all the details of the Compton camera such as Doppler energy broadening, detector energy resolution, detector segmentation, energy discrimination, etc. This study used the PENELOPE physics model in GEANT4 to accurately model the Compton scattering including atomic binding effect and Doppler energy broadening. The energy resolution of the scatterer and absorber detectors was simulated based on the measured data, but assuming ideal Gaussian distribution of the peak in the energy spectrum. The developed model simulates 20 keV and 10 keV energy discrimination level for the scatterer and absorber, respectively. Our result shows that segmentation of the detectors (especially, the 25-SEGD detector) significantly affects the spatial resolution of the Compton camera (FWHM = 1.7 cm for a point source). The Doppler energy broadening and detector energy resolution results in FWHM of 0.8 cm and 0.9 cm, respectively. The energy discrimination of the detectors was found to significantly affect both the sensitivity and spatial resolution. Our result suggests that a higher energy gamma source (e.g., 18F emitting 511 keV annihilation photons) should be used for Compton camera imaging. The use of higher energy gamma sources will significantly improve the spatial resolution of the Compton camera, nearly eliminating the effect of Doppler energy broadening and detector energy discrimination. It will also significantly reduce the effect of the detector energy resolution. The detector segmentation should be reduced down to a few mm or less to achieve the spatial resolution of 0.5 cm required in medical imaging. We believe that the 25-SEGD detector should be replaced with a more sophisticated detector such as a double- sided strip germanium detector or a stack of DSSD detectors.

Speaker: Ms So Hyun AN (Hanyang University)

14:00 Estimating accidental coincidences for pixelated PET detectors and singles list-mode acquisition

* Motivation: High spatial resolution and high sensitivity are mandatory in small animal PET. To achieve high resolution, pixelated crystal detectors with individual read-out are often employed. Sensitivity can be improved by reducing the scanner diameter, and also by lowering the low energy threshold (LET). The latter allows photons which undergo Compton interactions in the object or in the detector elements to be detected. A low LET also increases the number of detected accidental coincidences (randoms). To estimate the randoms rate, either a delayed coincidence window or the singles rate is used, although the data acquisition system usually constraints the choice. However, modern systems which are capable of acquiring singles list-mode data allow both techniques to be applied post-acquisition. This is the case for our small animal PET prototype MADPET-II, a high resolution scanner. * Objective: This work is aimed at studying the validity of random estimation techniques for low energy thresholds and singles list-mode data sets. We also investigate the role of inter-crystal scatter (ICS) in the estimation of accidental coincidences. * Methods: The scanner MADPET-II (under construction) currently consists of two opposing sectors, where each sector comprises three dual-layer detector modules. A layer consists of 4x8 LSO crystals coupled one-to-one to the channels of an APD matrix. The crystal surface is 2x2 mm2, with a thickness of 6 mm (front layer) or 8 mm (back layer). The scanner diameter is 71 mm. The present configuration has been simulated using GATE. Point sources were both measured and simulated, where the output in both cases was a singles list-mode data set. Additional simulations of sources of varying activity and geometry were also carried out. Coincidences were sorted post-acquisition in software using coincidence window widths from 5 ns to 60 ns, and LETs from 100 keV to 450 keV. The values of the time and energy resolution were 10 ns and 21% (at 511 keV), respectively. The number of random events was estimated (a) using a delayed coincidence window (DW method) and (b) the singles rate formula (SR method): Rij= 2 Tau Si Sj. The GATE file was also used to compute random events: A prompt coincidence was considered a GATE random if the two single photons detected within the coincidence window were related to different annihilations. * Results: For both simulated and measured data, the SR method predicted a higher number of random events than the DW method, and both estimation methods yielded more random events than the computed GATE randoms. This overestimation of randoms decreased with increasing LET values. The ratio R between estimated and GATE randoms, which accounts for the mismatch, depended on the source geometry, being smaller for extensive sources. On the other hand, R remained almost constant for coincidence windows larger than 10 ns. For LET=200 keV and for a centered planar source, R was approximately 1.6 for the SR method, and 1.15 for the DW method. For a simulated point source located outside the FOV (OFOV) and LET under 255 keV, the mismatch between the DW method and the computed GATE ranoms was R=1.12. We also observed that the number of prompt coincidences was larger than the number of GATE randoms. This was due to inter-crystal scatter coincidences, as the analysis of the simulated file showed. The contribution of ICS events to the promts was 27% for LET=150 keV and, as expected, diminished for higher LETs (10% for LET=200 keV). For LETs over 255 keV, the number of promts, GATE and DW random events was almost identical. The mismatch between DW and GATE randoms was less pronounced for this source location than for sources within the FOV. In the measurement of the OFOV source, scatter in the object also contributed to increase the number of prompts, so that the number of detected coincidences remained higher than the number of DW estimated randoms for all tested LETs. * Discussion and Conclusion: Our study shows the need to revisit the issue of random estimation when using pixelated crystals read out individually and low energy thresholds below 255 keV. The analysis of simulated data reveal that inter-crystal scatter plays a decisive role in the estimation of randoms, regardless of the method of choice. This is because pixellated detectors with individual crystal readout allow the single interactions related to a ICS event to be detected, if the LET is low enough. Moreover, since ICS single hits are related to the same annihilation, they may thus give rise to true coincidences. ICS can also cause more complex events to appear, such as those consisting of three single hits, two of them related through ICS to the same annihilation. The contribution of these complex events to the discrepancy between estimated and GATE randoms is currently under investigation. Our future work is also aimed at studying the role of energy resolution and object scatter in the estimation of random coincidences. For this purpose, further measurements and simulations will be carried out.

Speaker: Dr Magdalena Rafecas (Instituto de Fisica Corpuscular (IFIC), Universidad de Valencia/CSIC)

14:00 Evaluation of the conical scanning scheme for SPECT applications

Gamma-camera use for a breast imaging (Scinti-Mammography) is a quite difficult task because of specific geometry. Edge artefact is the main problem that should be solved in imaging system. Reduced FOV near edges relates to scintillation detector non uniformity and image reconstruction problems. The scale of the “dead zone” usually corresponds to the half of the PMT size used with detector plate. The non-planar conical scanning scheme is proposed to overcome the problem. The object is projected on the series of planes, which are tangent to some cone. The object is positioned inside cone aperture. In return for expanded field of view slight deterioration in spatial resolution, especially along cone axis was obtained. Whole set of projections doesn’t give complete data for the proper tomography reconstruction. The dependence of reconstruction accuracy on opening angle of the cone is analyzed and the mathematical simulation results are given for the several opening angles. Reconstruction algorithm was tested on phantom using conventional whole body gamma-camera. The obtained results prove the general concept. The use of proposed method allows extending FOV of SPECT systems, especially dedicated imagers such mammo-SPECT.

Speaker: Mr Vyacheslav Pedash (Institute for scintillation materials NAS of Ukraine)

14:00 EVENT MODEL SIMULATING IN OBJECT ANALYSIS SYSTEM

There are a lot of tasks in medicine, engineering and industry, where normal and critical states of some objects can be monitoring, recognized and registered as event signals in real time for analysis. Event recognition and registration of abnormal signals can be used to study of objects before and during its abnormal states or to predict its risk situations. Recognition and registration of real normal signals and abnormal events from objects in file server (data base) can be used for object diagnostics, analysis and to predict critical situation (knowledge base). Proposed Effective compact embedded modular system for signal registration can work as autonomous station or as network node. Distributed monitoring and registration system (MARS) with network-based architecture for event model simulating and analysis of complex objects are proposed and discussed. The Signal Measurement and Registration Terminal (SMART) node based on passive 3 PCI bus slots for SBC- and DSP- based modules with signal condition unit are described as example of industrial objects analysis. Typical or special events model simulating at the same inputs and the same instrumentations can be used for hardware testing, new software algorithms checking and more detail object events studying.

Speaker: Dr Vyacheslav Vinogradov (Institute for Nuclear Reserarch RAS)

14:00 Full 3D cluster based iterative image reconstruction tool for small animal PET camera

Iterative reconstruction methods are commonly used to obtain images with high resolution and good signal-to-noise ratio in nuclear imaging. The aim of this work was to develop a scalable, fast, cluster based, fully 3D iterative image reconstruction package for our small animal PET camera, the miniPET. The miniPET scanner consists of four detector blocks mounted on a rotatable gantry. Each detector block has 64 individual LSO crystal needles - with a size of 2 X 2 X 10 mm3 – arranged in an 8 X 8 array. The data characterizing the events is transmitted from the detector modules to a cluster of PCs by a communication module over an Ethernet network, using UDP/IP protocol. A Linux cluster of 12 PC-s is dedicated to data collection and image reconstruction. We have developed a software toolkit for implementation and comparison of different iterative algorithms. The toolkit is easily to adaptable to different detector geometries. The software kit contains both the general filtered back projection and also the ML- EM iterative image reconstruction method. The reconstruction package is developed to determine the 3D radioactivity distribution from list mode type of data sets and it can also simulate noise free projections of digital phantoms. An iterative image reconstruction process comprises of two parts, a detector geometry dependent one and another one dependent on the object under study. We separated these two parts in order to increase the calculation speed and the flexibility of the full 3D iterative reconstruction method. As the detector geometry is fixed for a given camera, the system matrix describing this geometry is calculated only once and used for every image reconstruction, making the process much faster. The calculated and stored system matrices represent the projection model of the camera: the number of detectors, number of detector rings, crystal geometry, detector position, gaps between the rings and gaps between the detectors. The pre-calculated system matrix together with any digital phantoms automatically ensure that noiseless simulated datasets can be obtained for a given detector arrangement. This is a reliable tool to compare simulated and real data sets. The Poisson and the random noise sensitivity of the ML-EM iterative algorithm were studied for our small animal PET system with the help of the simulation and reconstruction tool. The reconstruction tool has also been tested with data collected by the miniPET from a line and a cylinder shaped phantom and also a rat.

Speaker: Mr Ivan Valastyan (Institute of Nuclear Research of the Hungarian Academy of Sciences, Debrecen, Hungary Royal Institute of Technology, Stockholm, Sweden)

14:00 Gamma-ray Tracking Method Using a Digital Pulse Shape Analysis for Improving the Imaging Sensitivity of a Compton Camera

A Compton camera has excellent imaging and spectroscopic characteristics which can fulfill the requirements in various fields such as biomedical research, nuclear medicine, astrophysics, and national security. Especially its high imaging sensitivity is the most fascinating aspect to nuclear medical imaging fields. According to our previous works related to the performance of a proof-of-principle Compton camera based on a double-sided silicon strip detector (DSSD) and a 25- segmented germanium detector (25-SEGD), however, the imaging sensitivity was not sufficient to our expectation. The primary factor which makes it difficult to achieve the high imaging sensitivity is the rejection of multi-site events that are fired for two or more electrodes of the detectors due to multiple scattering of gamma rays. In order to improve the imaging sensitivity of our Compton camera by accepting the multi-site events in the image reconstruction, we have investigated a gamma-ray tracking method using a digital pulse shape analysis for the 25-SEGD. The method is to determine the most probable scattering sequence restricted by the Compton- scattering kinematics considering the interaction positions and energies for multi- site events. The interaction position and the energy of each scattering are determined by fitting a measured pulse shape with a calculated one by the weighting field method. We performed the test of our gamma-ray tracking method with a 137Cs (662 keV) standard gamma-ray source which was collimated by a lead block. The collimated gamma ray was incident on the center segment of the 25-SEGD. Pulse shapes of the central and three adjacent segments were measured by the DGF-4C pulse analyzer, manufactured by XIA. Experimental results will be compared to the Monte Carlo simulation. Contribution of the multi-site events to the improvement of the imaging sensitivity will be discussed.

Speaker: Mr Nam Young Kim (Department of Physics, Chung-Ang University, Seoul 156-756, Republic of Korea)

14:00 Geometric Sensitivity of ClearPET(TM) Neuro

ClearPET(TM) Neuro is a small animal PET scanner dedicated to brain studies on rats and primates. It belongs to the ClearPET family of small animal PET scanners that are developed within the Crystal Clear Collaboration (CERN) and use the same detector block design with LSO and LuYAP crystals in phoswich configuration, directly coupled with multi-anode photomultiplier tubes. ClearPET(TM) Neuro consists of 20 modules each with 4 detector blocks in line with 8x8x2 crystal matrices. Due to the extension of the photomultiplier tubes there are axial and transaxial gaps between the crystal blocks. To compensate for these gaps each second module is axially shifted and the scanner rotates during data acquisition. However, the design of ClearPET Neuro still leads to a specific geometric sensitivity, characterized by inhomogeneous and - depending on the measurement set-up - even incomplete sinogram data. With respect to reconstruction techniques, homogeneous and complete data sets are a 'must' for analytical reconstruction methods like Filtered Backprojection and the use of Fourier Rebinning, whereas iterative methods take the geometrical sensitivity into account during the reconstruction process. Nevertheless, also here a homogeneous as possible geometric sensitivity over the field of view is highly desirable. Therefore this contribution aims at studying the impact of different scanner geometries (axial shift, scanner radius) and different measurement set-ups (scanner rotation, various axial bed positions) on the geometric sensitivity. For that purpose a data set of coincident events is computed for certain settings that contains each possible crystal combination once. The lines or response are rebinned into normalizing sinograms and backprojected into sensitivity images using STIR (Software for Tomographic Image Reconstruction) tools. Both, normalizing sinograms and sensitivity images mirror the geometric sensitivity and therefore provide information which setting enables complete and homogeneous (as far as possible) data sets. An optimal measurement set-up and scanner geometry in terms of homogeneous geometric sensitivity is found by analyzing the sensitivity images.

Speaker: Dr Brigitte Gundlich (Forschungszentrum Jülich GmbH)

14:00 Image optimization and dose evaluation in coronary stenosis using multi slices computed tomography

Cardiovascular diseases are the most common incidence for premature death in developed countries. A major fraction is attributable to atherosclerotic coronary artery disease which may result in sudden cardiac failure. A reduction of mortality caused by myocardial infarction may be achieved if coronary atherosclerosis can be detected and treated at an early stage before symptoms occur. Therefore there is need for an effective tool that allows identification of patients at increased risk for future cardiac events. The current multi-detector CT has been widely used for detection and quantification of coronary calcifications as a sign of coronary atherosclerosis. The aim of this study is to optimize the diagnostic values and radiation exposure in coronary artery calcium screening exam using multi-slice CT (MSCT) with different image reconstruction algorithms. The phantom study is underwent routine MSCT coronary artery calcium scoring with retrospective ECG gated data acquisition. The radiation exposure for all protocols is evaluated by using computed tomography dose index (CTDI) phantom measurements. We found an optimal scanning protocol minimizing patient radiation exposure significantly to the MSCT coronary artery calcium screenings and still preserve its diagnostic accuracy compared with existing protocols by properly adjusting the CT scanning parameters and optimizing image processing techniques. These changes make the MSCT have more operation flexibility and provide more diagnostic values in current practice.

Speaker: Mr Yung-Hui Huang (National Yang-Ming University)

14:00 Inter-crystal scatter identification for a depth-sensitive detector using support vector machine for small animal PET

In a conventional PET detector, detected events are projected to 2D position histogram with Anger calculation for a crystal identification. However, the measured histogram is affected by inter-crystal scatterings (ICS) which occur in a whole detector. Peaks with projected each crystal in the histogram are blurred, and it causes ICS’s mispositioning. A depth of interaction (DOI) detector has been developed for the small animal PET scanner: jPET-RD. This DOI detector uses 32x32 crystals with four layers and a 256-channel multi-anode flat panel photomultiplier tube (FP-PMT) was developed by Hamamatsu Photonics K.K. Each crystal element is 1.45x1.45x4.5 mm. The FP-PMT has a large useful area (49x49 mm) and small anode pitch (3.04 mm). Therefore, the FP-PMT can extensively trace the behavior of incident gamma rays in the crystals including ICS. We therefore propose a novel method for ICS estimation using a statistical pattern recognition based on the support vector machine (SVM). In this study, we applied the SVM for discriminating photoelectric events from ICS events generated from multiple-anode outputs. The SVM was trained by uniform irradiation events generated from a detector simulator using Monte Carlo calculation. Initial result of ICS identification is about 80 % for non-training data. This method can archive a true subtraction of ICS from measured events, and it is also useful for random correction in PET.

Speaker: Dr Eiji Yoshida (National Institute of Radiological Sciences)

14:00 Investigation of a Whole-Body DOI-PET System

Recently, small animal depth of interaction (DOI) PET systems are being researched and developed in a variety of laboratories and companies. In this study, we were conducting basic research on a whole-body DOI-PET system which provides spatial resolution that is both high and uniform, and also minimizes costs. The detectors used consist of double-layer 9×10 GSO/GSO crystal blocks, a specially-designed light guide and two rectangular double anode type PMTs. Individual crystal sizes are 2.45×5.1×15mm3, and each layer of crystal blocks has a different decay time. Many of the circuit boards used in our current conventional PET system (SET-3000G SHIMADZU Japan) have been optimized for DOI acquisition. The detectors are arranged to form a 332.5mm radius detection ring, and spatial resolution is obtained from a center to 250mm radius field of view (FOV). The effect of DOI was confirmed using a comparison with the non-DOI system.

Speaker: Mr junichi ohi (shimadzu)

14:00 Investigation of the effect of the scintillator material on the overall X-ray detection system performance by application of analytical models

Computer simulation modeling has been proved as a reliable and promising technique for the evaluation of the performance of medical imaging systems. The purpose of the present work was to model modern X-ray detection system and to investigate the effect of scintillator material on the detector's output signal and the image quality. The scintillators were used in the form of screens (layers) of various thicknesses and the parameters investigated were the Modulation Transfer Function (MTF), the Detective Quantum Efficiency (DQE) and the Energy Absorption Efficiency (E.A.E.). The results of some commonly used scintillator detectors (YAG:Ce, YO3, ZnSCdS, Lu3Al5O3, CdWO4, LuO3) are presented. A computer algorithm, based on commercially available software, was developed to simulate the irradiation process. Typical radiographic conditions including the beam quality, in terms of the X-ray spectrum and exposure, were considered as input parameters (X-ray photons spectrum from 30keV to 140keV). For simulation purposes the intrinsic conversion efficiency (nc), the total number of optical photons produced per incident X-ray (m0), the attenuation coefficients and other optical parameters of the scintillator materials under study, were taken as input data. The complete simulation procedure was performed in a specially designed user-friendly Graphical User Interface (GUI). The results showed that, under typical examination conditions (80kV, 1mAs and 80 mg cm2 coating thickness), the MTF and DQE curves were affected by the X-ray energy absorption efficiency and the intrinsic properties of the scintillator materials. The Lu3Al5O7 scintillator presented similar behavior to that of ZnSCdS:Ag , exhibiting similar DQE at zero spatial frequency (0.16) and MTF at 200 lp (0.15). The YAG:Ce scintillator exhibited higher MTF at 200 lp (0.2) compared to the one of the YO3:Eu scintillator (0.1), but lower DQE at zero spatial frequency (0.062).

Speaker: Mr Nikolaos Efthimiou (Department of medical instrumentation-Technological educational insitution)

14:00 Investigations of the tomographic reconstruction issue in particular acquisition geometry in the framework of a bimodal microPET/CT

Biological research on small animals is constantly demanding for imaging devices with higher performances. Innovative micro-scanners are being developed in particular to enable combined acquisition of anatomical and molecular data. In this framework, we consider the idea of using the hybrid pixel detectors with the ClearPET (see poster " PIXSCAN: Pixel Detector CT-Scanner for Small Animal Imaging " at this conference) to built a prototype of microPET-CT which will permit simultaneous acquisition of the two modalities. The hybrid pixel detector XPAD used in the PIXSCAN microCT will be placed in front of the gamma detectors since it is expected not to interfere with the gamma rays of the PET. However the X-ray source will optimally be fastened to an independent gantry outside the PET ring which will involve a non-conventional off-centered circular geometry for the CT data acquisition. The purpose of this work was to investigate CT reconstruction with this geometry. Projection data collection along a circular source path is known to lead to approximate reconstruction due to missing projection data, except in the plane containing the source. No exact reconstruction can therefore be expected in off-centered geometry where the region of interest is shifted outside this mid plane. We first adapted the popular FDK algorithm to the reconstruction in off-centered geometry. Although our formulation slightly improves the results compared to that obtained with original FDK, strong artifacts remain when the off-center angle involved is important. These artifacts are well-known in reconstruction from data acquired along a circular source trajectory and are referred to as cone beam (CB) artifacts. We then evaluated several CB artifacts compensation methods as well as an algebraic reconstruction formula (SART) on off-centered data. If some attenuation of the reconstructed function can be corrected, strong geometrical deformations remain, in particular when the phantom used for simulation has strong gradients along the rotation axis.

Speaker: Mrs Solene Valton (CREATIS)

14:00 Measured and Simulated Specifications of the Lausanne ClearPET Scanner Demonstrator

Positron Emission Tomography (PET) applied to small animal imaging is a potentially important tool in developing new drugs and imaging gene expression. The Crystal Clear Collaboration (CCC) is developing a new family of small animal PET scanners called ClearPET, which are based on LSO/LuYAP Depth-Of-Interaction (DOI) sensitive detector modules. In parallel, GATE, a new Monte Carlo simulator based on Geant4 was developed to help in the design of the ClearPET prototypes. Measurements obtained with the partial ring Lausanne ClearPET scanner demonstrator are presented and compared against GATE simulations. For the present architecture, a maximum single event count rate of 1.34 Mcps is measured. This corresponds to a coincidence count rate of approximatively 31k cps. The random coincidences are estimated by delaying single events and the scattered ones by comparison with GATE simulations. Good agreements are observed between measured and simulated data. Count rate performance, including NEC curves, are determined and extrapolated for a full ring ClearPET design using GATE Monte Carlo simulations. A Mini-Derenzo phantom was also simulated with the aim to study the effect of depth-of-interaction (DOI) on image quality. Results are compared for different scanner designs corresponding to a full-ring ClearPET design with two crystal layers or only one crystal layer. As expected, image quality improves significantly with DOI both in terms of signal-to-noise ration and contrast.

Speaker: Martin Rey (LPHE, EPFL)

14:00 Monte Carlo simulation of the deposited dose in a patient during a therapy with carbon ions

Heavy charged particles like protons or light atom nuclei like carbon ions have interesting ballistic properties in cancer therapy. When used as projectiles, their electromagnetic interaction with matter is characterized with an energy loss spectrum peaking at a localized penetration depth: the so called Bragg peak. However, projectile particles used in hadrontherapy have sufficient kinetic energies to undergo nuclear reactions with target nuclei. As a consequence, depending on the nuclear process involved, multiple nuclear fragments can be produced. Even though the cross section of these processes would hardly exceed 40% of the total interaction cross section, they would contribute to a significant spread of the deposited dose over the target volume of interest. Despite this unwanted effect in hadrontherapy, among the produced radioactive isotopes a small fraction of those would β+ emitters and offers the only way for an online dose deposition control and even more the only way for the exploitation of inline and in- situ PET (Positron Emission Tomography) techniques for real time imaging. Here, we present our preliminary results concerning Monte Carlo GEANT 4 simulation to study the physical dose induced by fragments in a patient during an irradiation with carbon ions. The first step of this work concerns the improvement of the experimental database of the fragmentation’s cross sections in the range of energies used in therapy and their integration in GEANT4. After, we will present the proper implementation within Monte Carlo simulation framework of the patient data usually represented with a regular grid of heterogeneous voxels, obtained from a CT image. The use of this type of geometry in Monte Carlo requires the optimisation of the particle tracking to reduce the simulation time.

Speaker: Ms Nathalie DUFOUR (PhD Student - Centre Léon Bérard)

14:00 Monte Carlo validation in the diagnostic radiology conditions.

Radiation transport phenomena have been extensively studied by application of the Monte Carlo technique. This was proven to be by far the most successful technique for the simulation of the stochastic processes involved in radiation detection. During the last decade, various Monte Carlo simulation packages have become commercially available, however constrained by expediency and feasibility. One of these, MCNP, is a general-purpose, generalized geometry, coupled photon/electron/neutron Monte Carlo transport code. The present paper aimed to the validation of Monte Carlo simulation codes already developed by the reporting team, for the study of photon transport, photon absorption and x-ray fluorescence generation phenomena occurring in scintillators employed, in ordinary x-ray medical imaging modalities (general conventional and digital radiography-fluoroscopy and computed tomography).Comparisons are reported between the developed codes, MCNP developed codes and other published data. First, the depth of energy deposition in water was assessed for three monoenergetic x-ray beams (15, 20, 30 keV). A water slab of 10 cm thickness and infinite width was simulated. Exposure was modeled as a narrow beam of photons normally impinging. The energy deposited in slabs of varying depths was tallied. Excellent agreement within ±2% was achieved, except that for 15keV were a more rapid drop with increasing depth was found. Second, the lateral spread of energy deposition was assessed in a 1cm thick slab in the centre of an 8cm thick water phantom, irradiated by a 50keV narrow beam. Again, agreement within ±2% was achieved. Third, a water slab with thickness 5, 10, 15, 20 cm was modeled irradiated by a monoenergetic narrow beam of photons of various energies. The mean number of interactions for each incident photon was determined and compared to published data. Agreement to within ±2% was achieved. Last, the relative scattered x-ray photon fluence as a function of exit angle was determined, using a 4cm thick Plexiglas phantom irradiated by a 27.3keV photon beam. Aggreement between published and modeled data was also found. Further comparisons were performed with published data on Gd2O2S scintillator (coating thickness of 90 mg/cm2 at disk geometry) exposed to a photon source without collimation 10cm above the disk.

Speaker: Dr Dimitrios Nikolopoulos (Department of medical instrumentation-Tchnological educational insitution)

14:00 MR Image Reconstruction from Partial k-Space Using Singularity Function Representation

Partial k-space acquisition is usually employed in magnetic resonance imaging (MRI) for reducing imaging time while maintaining image quality. In this field, image reconstruction from the incomplete k-space is an important matter. A number of reconstruction methods have been reported in the literature, but all these methods do not cope with truncation artifacts. We present a method that is particularly suitable for reconstructing magnetic resonance (MR) images from partial k-space by reducing substantially truncation artifacts. The proposed method is based on the use of a so-called singularity function representation. It consists in first representing the image to be reconstructed by the linear combination of singularity functions (step functions), and then estimating the parameters of the representation model through using information derived from the truncated k-space. In contrast to MR reconstruction using, directly or indirectly, inverse Fourier transform, this method does not reconstruct images directly from k-space. The key point of this method using singularity function-based reconstruction is the appropriate determination of singular points and singularity degrees. Therefore, we present a new strategy for estimating these two parameters by imposing constraints. It consists of first extracting singular points using a layer extraction technique, and second using the obtained singular points in the equation system for obtaining singularity degrees. We restrict the singularity degrees within some dynamic range, beyond which the corresponding singular points are set to zero. This new approach allows to reduce or suppress false singular points, and thus improves the reconstruction quality. The proposed MR reconstruction method was evaluated on both simulated and acquired MR data from human brain. For simulated data, a slowly varying phase was introduced to generate a complex image whose k-space does not exhibit Hermitian symmetry. The acquired MR data saved as raw k-space data were obtained using a 1.5 T Siemens Sonata system. The reconstruction quality was assessed visually and using the quantitative criteria such as normalized mean-square error (NMSE). The obtained results showed that the method is particularly efficient for overcoming reconstruction limitations due to truncation artifacts in partial k-space acquisitions. It exhibits a substantially improved performance in comparison with the classical and popular zero-filling technique and more state of the art methods such as the Margosian/Homodyne method. In conclusion, the proposed method allows to obtain images of good quality with a significant reduction of scanning time by maximizing the asymmetry of k-space.

Speaker: Dr Yuemin ZHU (CREATIS, CNRS UMR5515 & INSERM U630, Lyon)

14:00 New reconstruction algorithm for directional-derivative projections of the high-resolution DEI-CT

X-ray diffraction enhanced imaging (DEI) has extremely high sensitivity and resolution of weakly absorbing low-Z samples, so it becomes an effective method for molecular imaging recently, especially in the medical and biological fields. Many experiments have been performed and high-resolution phase-contrast images of many elaborate details of the samples, such as human breast cancers and joint cartilages et al, have been obtained using the DEI method. The computed tomography of the DEI (DEI-CT) can reconstructed the distribution of the refractive index decrement (n') or the refractive index (n) of the sample, where n'=1-n. But the basic theories and reconstruction algorithms are different from traditional absorption-contrast CT if the rotated axis of the sample is parallel to the diffractive plane of the analyzer. Here, the projections of this CT mode are called ‘directional-derivative projections’. This reconstruction problem of directional-derivative projections have been investigated by K.M.Pavlov, I.Koyama, A.Maksimenko and Pei-Ping Zhu et al, who proposed different ‘two-step’ algorithms, including ‘restoration then reconstruction’ and ‘reconstruction then restoration’ methods. However, the restoration process may result in artifacts and inaccurate values in CT images. This paper is dedicated to a direct reconstruction algorithm for directional- derivative projections. The new algorithm can directly reconstruct the distribution of the refractive index decrement of the sample without restoration process like the filtered backprojection algorithm (FBP) of the traditional CT. A mathematical deduction is described in detail, based on the differential theory of two- dimensional Fourier transform and the theory of polar coordinate in the frequency domain. A reconstruction formula of directional-derivative projections similar to the basic formula of the FBP algorithm is obtained. The difference of the new algorithm from the FBP algorithm is that it uses the Hilbert filter instead of the RL or SL filter in the frequency domain. It can obtain accurate values of the refractive index decrement of the sample without any calibration. Finally, a computer simulation of the DEI-CT is presented to produce directional-derivative projections. The simulated cylinder is rotated in a step of 1.0 degree within 180 degrees, so 180 directional-derivative projections in total are obtained during the simulated experiment. The projections are with or without noise. For projection in each direction, 41 DEI images are measured in different positions of the simulated rocking curve. Afterwards, refraction-angle images are calculated by the multiple- image statistical method. The results from the new algorithm are compared with the results from ‘two-step’ algorithms using with-noise and without-noise projections. The new algorithm is well validated in the experiment. Since it is a ‘one-step’ process and does not require any restoration process, it avoids artifacts brought by restoration, especially for noisy projections. In conclusion, the new algorithm makes an advance in this field. We believe that it is a perfect form of the direct computed tomographic reconstruction algorithm for directional-derivative projections of the DEI-CT.

Speaker: Dr Zhi-Feng Huang (Dept. of Engineering Physics, Tsinghua Univ., China)

14:00 Non Contact Fluorescence Optical Tomography by means of Numerical and Analytical Approaches

Near-infrared fluorescence-enhanced Diffuse Optical Tomography (fDOT) has been proven to be an efficient tool for image reconstruction of the bio-distribution of fluorescent markers. By labeling regions of interest such as tumors with target- specific fluorescing molecular probes, fDOT enables both the three-dimensional (3D) localization of the targeted areas and the quantization of the local concentration of the fluorochromes. Within this framework, the reconstruction algorithms use either analytical or numerical forward solvers that are required to be fast as well as robust. Analytical approaches are time efficient but are usually restricted to the examination of objects with simple geometries (infinite, semi-infinite media, slabs, cylinders…), leading to the necessity of using an index matching fluid for immersion of the animal in order to satisfy this assumption; whereas numerical ones, such as the Finite Element Method (FEM), are more versatile but are known to be time and memory consuming. The present work is motivated by the fact that one of the requirements for small animal imaging is to get rid of the constraint of the index matching fluid and to image the animal in free space. Here, a comparison between two different approaches, analytical and numerical, for the establishment of an efficient forward solver for using fDOT technique in a non contact geometry is presented. Both are adapted to the use of the CCD technology for the detection. The experiments have been performed with our laboratory made tomographer. The optical system is composed of a laser source (690 nm, 8 mW) for illumination and a CCD camera (Orca ER, Hamamatsu) for detection. The source is guided to the object via an optical fiber. The movements of the scanning fiber are driven by two translating plates (Microcontrol) and monitored with a computer. The CCD camera is focused at the top surface of the object. For fluorescence light detection, a filter (high pass RG9, Schott) is placed in front of the camera. A 2.5cm wide half cylindrical solid phantom has been designed for the experiment. It is composed with a mixture of epoxy resin (Solloplast), titanium dioxide powder (Sigma-Aldrich) as the scatterers, and black ink as the absorber. The index of refraction is estimated to be 1.54, the diffusion coefficient 10 cm-1, and the absorption coefficient 0.2 cm-1. In order to easily introduce fluorescent inclusions, four holes have been drilled, at different positions. To model the fluorescent regions, we considered two thin glass tubes (external diameter: 3 mm, internal diameter: 1 mm, length: 3 cm) filled with commercial fluorescent dyes (Alexa750, 10 microM, Molecular Probes), introduced in the phantom at 12 mm from the ground surface. Both are separated by a distance of 2 cm. In this experiment, 12×9 different sources positions, equally spaced, and separated by a distance of 3 mm have been considered, corresponding to a rectangular field of view of approximately 3.3×2.5 cm2. The numerical approach is based on the resolution of both the forward and the adjoint problems by using the FEM. A segmentation-based method is adopted for constraining the reconstructions with a classical ART algorithm. In the present analytical treatment, the morphology of the object is determined indirectly by an appropriate reconstruction of optical heterogeneities. As a consequence, this accounts self-consistently for the errors produced by using an analytical model. An example of reconstructions obtained by using the two kind of forward solvers was performed. There are no obvious differences between the results obtained with one approach or the other. An even more surprising remark, if a close-up is made on these results, is that those performed by using the analytical forward solver are slightly better. This is due to the fact that this approach takes into account most of the causes of the discrepancies between the actual object and the assumptions on the geometry or on the homogeneity made. It accounts efficiently not only for the morphology of the object but also for the possible presence of heterogeneities (in this case the tubes) inside the object. A complete description of the models and a precise comparison between the two approaches will be presented at the conference. We developed two viable techniques which address our goal of suppressing the matching fluid to perform fluorescence tomography on small animal.

Speaker: Dr Lionel Hervé (LETI - CEA Recherche Technologique)

14:00 PENELOPET, A monte carlo-based application for tomography based on PENELOPE

PENELOPE is a code for Monte Carlo simulations of the transport in matter of electrons, positrons and photons with energies from a few hundred of eV to 1 GeV. It is robust, fast and very accurate, but it may be unfriendly for people not acquainted with the fortran prograaming language. We have developed an easy to run application that allows complete simulations for PET and SPECT with PENELOPE. Very sophisticated simulations can be prepared by modifying just a few simple inputs files. The output data gets ready for post-analisys with differents levels of post-processing and can be analized with the prefered programming language. Parameters of the input files are scanner geometry and composition. The same for the objects to scan, source activity and isotope. A number of options, such as simulation of positron range, photons non-collinearity and scanner motion can be easily selected from the input file. It has also been implemented the posibility of limiting the number and kind of particles involved in the simulation. For instance, it can be choosen to simulate only the initial gamma photons (511 keV and others). The output files can have three levels of post-proccesing. In the lowest one, all the information about each interaction is kept for further analysis. At an intermediate level, just the singles events with the information needed for the analysis is stored. The possibility of pile up and cross talk are taken into acount. The third and highest level of proccessing stores the coincidence counts in a typical LIST file but with information about pile up, scatter, randoms and autocoincidences events from the simulation. A more elaborated analysis is possible if the user writes his own code. We get better performance with this applicattion that with other SPECT and PET dedicated codes such as SIMSET and GATE. Accurate simulations results in a reduced period of time are obtained. We made realistic simulations to validate the code against an actual continuous rotating PET scanner and a full ring PET scanner. Detector efficiency, sensitivity, spatial resolution, scatter and random fraction and energy spectrum are some of the parameters that we have measured to complete the validation. We can conclude that this is a good application for PET and SPECT simulations.

Speakers: Ms Esther Vicente Torrico (Hopital General Universitario Gregorio Marañón), Mr Joaquín López Herraiz (Universidad Complutense de Madrid), Mr Samuel España Palomares (Universidad Complutense de Madrid)

14:00 Performance characteristics of a cluster based data acquisition system dedicated to PET scanners

Practical engineering aspects of the next generation nuclear medicine detectors with Ethernet based data transfer entrust the solution of high speed data acquisition, real time event processing and data storing to software developers. In response to this challenge, a new cluster based data acquisition library (DAQ) was elaborated as part of a complex, object oriented, platform independent medical imaging software development environment. The communication between individual components (nodes) of our data acquisition system and the processing of the detector events are completely separate. Intranodal communication is achieved by traditional multithread server-client relations using standard stream and datagram sockets. A special message parsing protocol ensures the control, manipulation and monitoring of the nodes of the data acquisition network. The separation of the communication layer from event processing is a substantial advantage of this library acquitting the application developer from thread- and socket manipulations leaving them for the DAQ-developer and maintenance team. This performance test is aimed at (1) the comparison of the bandwidth of DAQ-protocol based data transfer to the theoretical one, (2) the determination of bandwidth and computing capacity demand required by the communication layer, (3) the analysis of the dependence of the effective data bandwidth on the number of detectors and the type of the applied real-time event processing algorithms, and (4) the hardware requirement (in terms of number of type of processors and network parameters) estimation of small animal positron emission tomographs of various detector configuration. Shortage in Ethernet based real PET detector systems capable of generating an arbitrary number of detector events of different count rates explains why only detector emulators were used in this study. Performance data were obtained using a special double-backboned Linux-cluster of 7 Supermicro hosts with dual 64-bit 3.0GHz Xeon processors and dual Gigabit Ethernet interfaces. Hosts were connected to both the local area network and the gigabit data acquisition network. Our data show that the bandwidth and calculation capacity demand of the communication layer is negligible. A calculation method is proposed to estimate the hardware requirements of any small animal or human PET scanner controlled by the outlined type of software.

Speaker: Dr Miklos Emri (PET Center, University of Debrecen)

14:00 Performance evaluation of a Small Animal PET Scanner. Spatial resolution characterization using 18-F and 11-C

Introduction The main objective of this work was to characterize the physical performance of a small animal PET scanner (GE eXplore Vista). The detector modules of the eXplore Vista consists of an 13 x 13 array of 1.45 mm square “phoswich” elements providing depth-of-interaction information. Each element is made of two scintillator crystal layers of Lutetium Yttrium Orthosilicate (LYSO) and Gadolinium Orthosilicate (GSO) optically glued together in order to form a single 15 mm long segment. The 36 detectors blocks are arranged in two rings giving a total number of phoswich elements equal to 6084. The transverse field of view (FOV) is equal to 6 cm while the axial FOV is equal to 4.6 cm. Typically PET images of small animals are performed using 18-F and 11-C labelled tracers. Because the mean positron range for the two isotopes is respectively equal to 0.6 mm and 1.1 mm, it is interesting to investigate the effect of such difference on the scanner spatial resolution. Materials and methods In order to evaluate the performance of the eXplore Vista several parameters such as spatial resolution, sensitivity, scatter fraction (SF) and noise equivalent counts (NEC) were measured. Spatial resolution was measured by using a point source, a line source and a custom made high resolution phantom. The point source (about 0.2 mm diameter) was obtained by soaking a small piece of zeolite with a 18-F or 11-C solution. The line source was obtained by filling a small needle (internal diameter of 0.3 mm) with 18-F or 11-C isotopes. In order to evaluate the dependence of spatial resolution with respect to transaxial axis position, the line source filled with 18-F was moved along the y axis and four measurements spaced by 5 mm were acquired. Images of a custom made high resolution phantom (HRP) with rectangular arrays of holes of 2, 1.5, 1.0 and 0.5 mm with the centres spaced by 4, 3, 2, 1 mm were also acquired. The HRP was filled with a solution of about 15 MBq of 18-F or 11-C and images were acquired for 1 h. All the images were reconstructed using Fourier rebinning and filtered back projection (ramp filter). In order to measure the absolute eXplore Vista sensitivity a small 18-F point source of about 1.2 MBq was placed at the centre of the field of view (FOV). The SF measurements were carried out using a small capillary (0.8 mm internal diameter) in air and in water (3 cm diameter cylindrical phantom). The SF measurement was performed accordingly with a method similar as the one described in the NEMA NU2–2001 report. In order to estimate the NEC values a rat sized phantom and a mouse sized phantom were filled with an high amount of 18-F (about 200 MBq) and let to decay for six half life. Measurements were acquired every 15 minutes. Results All the results described in this section are referred to the 250-700 keV energy window. The transaxial full width half maximum (FWHM) with the point source at the centre of FOV are respectively equal to 1.48 mm and 1.74 mm for 18-F and 11-C. The 18-F axial resolution is equal to 1.77 mm giving a volumetric resolution of 3.82 mm^3, while for 11-C the axial resolution is 1.91mm and the resulting volumetric resolution is equal to 5.78 mm^3. The FWHM measured with the 18-F line source at four transaxial positions were equal to 1.47 mm, 1.71 mm and 2.1 mm at the centre, 1 cm and 2 cm off-axis, respectively. Analysis obtained by drawing several profiles along the HRP hot spots confirmed a worsening in spatial resolution when using 11-C instead of 18-F (for high resolution small animal scanners). The absolute eXplore Vista sensitivity was equal to 2.2% (400-700 keV energy window), 3.9% (250-700 keV) and 5.9% (100-700 keV) without dead time correction applied. For a line source, the measured SF value in air was equal to 19.3%. SF values obtained in water were equal to 26.2% and 34% for a mouse and a rat-size phantom. The NEC peak rate were respectively equal to 147 kcps at 710 kBq/ml and 83 kcps at 169 kBq/ml for a mouse and rat-size phantom. Conclusions Results show that the eXplore Vista provides high spatial resolution and at the same time good sensitivity. The value of the spatial resolution for 11-C labelled tracers was worse respect to 18-F, such loss of resolution for 11-C was also evident by looking at in vivo mouse images. The eXplore Vista has been in use at our institution for the last twelve months showing good stability and high throughput.

Speaker: Dr Antonello Spinelli (Policlinico S. Orsola-Malpighi, Bologna, Italy)

14:00 Performance of a Small Animal PET Scanner Based on Monolithic Scintillation Detectors

In recent years, small-animal positron emission tomography (PET) has gained much interest. Most current designs use small scintillation crystals coupled to position- sensitive photomultiplier tubes. It would seem relatively straightforward to increase the resolution in such designs by decreasing the pixel size. However, this results in loss of sensitivity because of the increased dead space between the pixels. Also, resolution improvement may be impeded by inter-crystal scatter and parallax errors. Monolithic scintillation detectors, consisting of a few cm3 of scintillating material coupled to one or more position-sensitive light sensors, in particular avalanche photodiode (APD) arrays, can avoid these problems. In such detectors, the entry point of an incoming annihilation photon can be estimated from the distribution of the scintillation light on the APD arrays. Using the Monte Carlo code GATE, a small-animal PET scanner based on monolithic scintillation detectors is simulated in order to investigate the resolution of the reconstructed images and the count rate performance. The scanner consists of four rings of 38 detector modules each, with an inner diameter of ~128 mm. The modules consist of 20x10x20 mm3 LSO crystals coupled to two Hamamatsu S8550 APD arrays. Crystals with a trapezoidal rather than a rectangular shape are also investigated. By increasing the width of the crystal on the outside of the scanner, the dead space between the detector modules is minimized allowing for a further increase of the detection efficiency. These monolithic designs are compared to a scanner with the same dimensions using pixellated crystals. We have previously shown that detectors based on monolithic scintillators have an intrinsic spatial resolution similar to that of scintillation detectors using pixellated crystals (< 2 mm FWHM). Reconstructions showing the high resolution achieved in actual PET images can be simulated in a realistic matter using experimentally determined error distributions. These error distribution are obtained as a function of the photon entry point and the incidence angle by determining the difference between the true entry point and the measured one for a large number of events. Images and detailed spatial resolution results will be shown at the conference. The monolithic scintillation detector appears to offer a point-source sensitivity that is roughly a factor of two larger than a pixellated one. The same appears to be true for the noise equivalent count rate (NECR) for activities in the range of interest for small-animal PET (less than ~50 MBq), even at a relatively long dead time of 1000 ns. NECR curves calculated for a cylindrical water-filled phantom for different dead times, coincidence windows and energy windows will be presented at the conference.

Speaker: Dr Dennis R. Schaart (Delft University of Technology)

14:00 PET Image Reconstruction: A Stopping Criterion Based on the Updating Coefficients of the ML-EM Algorithm

We have studied certain properties of the maximum likelihood expectation maximization (ML-EM) algorithm for iterative image reconstruction in positron emission tomography (PET). The principal aim of the work has been the development and evaluation a new stopping criterion for this algorithm. We developed and used a software platform based on Monte-Carlo techniques, which simulates the emission of gamma rays in the source and their detection in the tomograph, to compute the transition matrix and also to generate projection data for reconstruction. The digital Hoffman brain phantom was used. A single-ring tomograph has been simulated with 128 detector crystals on the ring, which has 15 cm radius and a field of view (FOV) of 20x20 cm2. The images were reconstructed over a 64x64 and 128x128 grid. In order to investigate the problem of the deterioration of the image quality after a number of iterations, we have studied the statistical properties of the updating coefficients in the ML-EM algorithm. The results of this study show that the values of the updating coefficients for the non-zero reconstructed pixels follow a distribution composed of a peak region around 1 which becomes progressively narrower as the iteration proceeds, and also a tail below 1, which corresponds to that part of the image that is far from being completely reconstructed. This tail has an exponential form and is pushed to high values as the iterations progress. We find that the minimum and average values of the tail are related to the quality of the image. In particular for optimally reconstructed image these values are independent of the number of counts and of the image size. The possibility of exploring this property in stopping the algorithm at optimally reconstructed image is discussed.

Speaker: Dr Pedro Guerra

14:00 Simulation studies of an inline PET prototype

Heavy charged particles like protons or light atom nuclei like carbon ions have interesting ballistic properties in cancer therapy. When used as projectiles, their electromagnetic interaction with matter is characterized with an energy loss spectrum peaking at a localized penetration depth: the so called Bragg peak. This feature, combined with a raster scanned beam modulated in energy and intensity, is used in hadrontherapy to deliver volumetric dose distributions that conform to the physician prescriptions. Thus, it optimizes the energy deposition in the target volume (tumor) and spears the surrounding media. However, projectile particles used in hadrontherapy have sufficient kinetic energies to undergo nuclear reactions with target nuclei. As a consequence, depending on the nuclear process involved, multiple nuclear fragments can be produced. Even though the cross section of these processes would hardly exceed 40% of the total interaction cross section, they would contribute to a significant spread of the deposited dose over the target volume of interest. Despite this unwanted effect in hadrontherapy, among the produced radioactive isotopes, a small fraction of those is β+ emitter and offers the only way for an online dose deposition control and even more the only way for the exploitation of inline and in-situ PET (Positron Emission Tomography) techniques for real time imaging. Here, we present our preliminary results concerning simulation studies of an inline PET prototype using the dose dependent predicted β+ emitters distributions for a real time control of a treatment plan in hadrontherapy.

Speaker: Ms Marie-Charlotte RICOL (Institut de physique nucléaire de Lyon (IPNL) - Université Claude Bernard Lyon 1 (UCBL))

14:00 Study of randoms-induced artefacts in several small animal PET scanners

The design of current small animal PET tomographs, with open geometries and extended field of view (FOV), tries to optimize sensitivity, what makes them to be particularly prone to random coincidences. Particularly in 3D mode, random counts are an important contribution to measured events. The effect of accidental coincidences in the noise equivalent counts (NEC) of an acquisition is well known, but the appearance of artifacts in the images due to them is not so well established. For instance, we have observed that block-based scanners may show streaks in images recontructed from uncorrected data, corresponding to the edges of the detectors. We have investigated the artifacts induced from random coincidences in real PET scanners of different geometries. Some scanner configurations are shown to be particularly sensitive to these effects, and the quality of the images from high activity sources, is clearly deteriorated by these artifacts. These results show the importance of developing accurate methods to estimate random contributions and of correcting the measured signal or the system model, not only in order to increase the signal-to-noise ratio or to obtain quantitative images, but to avoid undesirable artifacts as well. This is especially important in some PET scanners configurations. We present the effect of random coincidences for different simulated scanner designs, and how the artifacts introduced depend on the scanner geometry, the shielding of the detectors and the ratio between FOV and detector size. Some designs optimized to reduce random-induced artifacts are proposed.

Speaker: Mr JOAQUIN LOPEZ HERRAIZ (UNIVERSIDAD COMPLUTENSE DE MADRID)

14:00 The Influence of Noise in Full Monte Carlo ML-EM and Dual Matrix Reconstructions in Positron Emission Tomography

Monte Carlo (MC) simulations in positron emission tomography (PET) play an important role in detector modeling and algorithm testing. Whereas the simulations are widely used in a forward projection manner to accomplish this task, ideally they should be included into the reconstruction process itself. It is therefore desirable to investigate the convergence properties and the propagation of MC noise of these kind of reconstruction algorithms. For human scanners the correct treatment of patient scatter plays a dominant role. The incorporation of this kind of scatter into the matrix is therefore important. MC simulations were integrated into the maximum likelihood expectation maximization(ML-EM) algorithm in two different ways. In the full matrix approach the system matrix was calculated by running MC simulations including scatter. This matrix was used in both the projector and the back-projector. In the dual matrix (DM) approach, MC simulations were used to incorporate scatter in the projector, whereas the back-projector only comprised attenuation. Repeated reconstructions with different MC seeds allowed a statistical analysis of the error at each iteration step and made it possible to investigate separately the propagation of the MC noise that was introduced by the sinogram, by the projector, and by the matrix. Both approaches resulted in similar images, but the DM approach with unmatched projector and back-projector yielded a faster initial convergence and a faster divergence at higher iteration numbers when compared to the ideal full matrix approach. The analysis of the noise sources for the modeled 2D-scanner in full matrix reconstruction showed that the noise introduced by the matrix became comparable to the noise introduced by the sinogram when using a matrix that was simulated with 10000 emissions/voxel or less using variance reduction techniques. The time needed to simulate this matrix with 6400 voxels and 234700800 elements was less than four minutes on a small computer cluster with eight two-processor computers.

Speaker: Mr Niklas Rehfeld (Section for Biomedical Physics, Radiooncology, University of Tuebingen)

14:00 Validation of a GATE model for the simulation of the Siemens PET/CT biograph™ 6 scanner

The recently developed GATE, a Geant4 Application for Tomographic Emission, is a Monte Carlo simulation platform developed for PET and SPECT and is freely distributed by the OpenGATE collaboration. GATE is based on a powerful simulation core, Geant4 toolkit, that includes well-validated physics models, geometry modeling tools and efficient visualization utilities, and, uses newly developed software components for data analysis and image reconstruction. GATE provides the ability of modeling time-dependent phenomena, such as geometry element movements and source decay kinetics, allowing the simulation of time curves under realistic acquisition conditions. The purpose of this paper was to validate a GATE model for the simulation of the Siemens PET/CT biograph™ 6 scanner. This three-dimensional GATE model simulated 24336 LSO ( Lutecium Oxyorthosilicate) detectors grouped into 144 blocks as the vendor’s specifications. The GATE results were compared to experimental data obtained in accordance to the NEMA NU 2-1994 and NEMA NU 2-2001 protocols. The Phantoms used by these protocols were also modeled within GATE, allowing us to simulate the sensitivity, scatter fraction, count rate performance and spatial resolution. The simulated sensitivity was within 1% compared to the experimental one, the simulated scatter fraction agreed within 2.5% of the measured value. The experimental peak true count rate and the peak activity concentration were matched by the simulated results to within 0.5% and 0.2% respectively. The spatial resolution of the simulated scanner matched the experimental results to within 0.4 mm.

Speaker: Mr Panagiotis Gonias (Department of Medical Physics, School of Medicine, university of Patras,Patra,Greece)

14:00 Validation on an Anthropomorphic Phantom of FORE Optimization in 3D PET

INTRODUCTION: Fourier Rebinning (FORE) [1] is the most widely used algorithm for the rebinning of 3D PET data into a stack of 2D sinograms. FORE operates into the sinogram Fourier Transform domain and it is based on the frequency-distance relation. The validity of this approximation holds at high frequencies; low angular, k, and radial, ω, frequencies are instead usually rebinned with Single Slice Rebinning (SSRB) limited to low copolar angle data. A square low frequency region is commonly defined on empirical basis and implies an abrupt transition of rebinning strategy. In a previous work [2] we proposed a simple index able to map the validity of the frequency-distance relation in the sinogram frequency-space. A gradual rather than an abrupt transition mask was consequently proposed for an optimized FORE. Preliminary results on scanned line and uniform phantoms showed the superior performances of the proposed method with respect to the standard abrupt transition: ring artifacts on reconstructed images were eliminated and a better compromise between axial resolution and noise resulted achievable. An anthropomorphic phantom allows to test the method over more complex structures in a condition more similar to oncological studies. METHODS: We compared the performances of standard abrupt and gradual partitions on the Alderson thorax-abdomen phantom, scanned with ECAT EXACT HR+. The phantom districts were filled with different activities simulating the in vivo 18F-FDG uptake and spherical lesions of different radii (from 2.15mm up to 10.96mm) were positioned inside. Spherical Volumes of Interest (VOIs) were defined on the spheres (80% of the sphere radius) and on both thorax and liver (radius 16mm). Mean imaged activity normalised to the total activity of the reconstructed study was extracted within each VOI. Inside the liver and thorax VOIs the std/mean ratio was also evaluated. Two successive acquisitions were performed and the obtained results were averaged. Both abrupt and gradual partitions were implemented with different parameters corresponding to a different extension of the low frequency SSRB region: for the first approach square SSRB regions were considered starting from (ωlim=0, klim=0) up to (ωlim=6Δω, klim=6); for the second one, in setting the gradual transition border, several values T of the validity index were tested starting from 1% up to 13% of the maximum validity level. RESULTS: When FORE was applied at very low frequencies, in a low validity range, the axial spread amount leaded to an important partial volume effect. For example for the 7.83mm sphere the imaged activity was 79% of the true activity for (ωlim=0, klim=0) and 84% for T=1%. The inaccurate rebinning of the low frequencies also caused an incorrect activity definition on large organs, i.e. 130% of the true activity inside the liver VOI for (ωlim≤Δω, klim≤1) and T≤3%. When the transition border was moved towards higher frequencies the axial spread was reduced: for example the imaged activity on the 7.83mm sphere was ≥92% of the true activity for (ωlim≥3Δω, klim≥3) and T≥3%. Although, a larger parameter choice implies a count statistics reduction, limited effects were observed with the proposed gradual partition; e.g. in the liver VOI the std/mean was ≤0.36 for all the explored parameter T values. The abrupt method, on the contrary, displayed std/mean ≥0.41 for (ωlim≥3Δω, klim≥3). CONCLUSION: Results previously observed on line and uniform phantoms relevant to minimization of axial blurring and artefacts were confirmed and were not jeopardized by the acquisition of more complex structures. We can conclude that the proposed method, by eliminating ring artifacts and reducing low frequencies noise component, allows a parameter choice corresponding to a better quantification accuracy and can be applied in clinical conditions. REFERENCES: [1]: M. Defrise, P. E. Kinahan, D. W. Townsend, C. Michel, M. Sibomana and D. Newport, “Exact and approximate rebinning algorithms for 3D PET data,” IEEE Trans. Med. Imag., vol. 16, pp. 145–158, Apr. 1997. [2]: E. De Bernardi, M. Mazzoli, F. Zito and G. Baselli, “Evaluation of frequency- distance relation validity for FORE optimization in 3D PET,” submitted for publications to IEEE Trans. Nucl. Sci.

Speaker: Dr Elisabetta De Bernardi (Bioengineering Department, Polytechnic University of Milan, Milan, Italy)

15:00 → 18:30 Tutorial

15:00 New Trends on Photodetectors

The working horse for the detection of photons is the photomultiplier tube (PMT) which was invented in the RCA laboratories and became 1936 a commercial product. It is an elaborated device but still, after 70 years, impressing improvements have been achieved recently. PMTs however have two severe handicaps: they are very sensitive to magnetic fields and their price is high because the complicated mechanical structure inside the vacuum container is mostly handmade. This forced the search for alternatives. Semiconductor devices, PIN-photodiodes, avalanche photodiodes and recently Geiger-mode avalanche photodiodes have been developed and have already replaced PMTs in many fields of research and will gain more ground in the near future. Specially Geiger-mode avalanche photodiodes have a high potential because they have high gain and need no or only a simple amplifier and they can be produced in a standard and cost effective CMOS technique. When very large areas have to be covered the cheapest device is probably a gaseous photomultiplier. Detectors based on the so-called GEM foils (Gas Electron Multiplier), which allow to minimize the harmful ion feedback, have high gain and even so a very long lifetime.

Speaker: Dr Dieter Renker (Paul Scherrer Institute)

15:30 The Road to Time-of-Flight PET

Simple theory predicts that the signal to noise ratio in PET can be reduced by an order of magnitude by using time-of-flight (TOF) information. This reduction can be obtained by improving the coincidence timing resolution, and so would be achievable in clinical, whole body studies using with PET systems that differ little from existing cameras. The potential impact of this development is large, especially for oncology studies in large patients, where it is sorely needed. TOF PET was extensively studied in the 1980’s but died away in the 1990’s, as it was impossible to reliably achieve sufficient timing resolution without sacrificing other important PET performance aspects, such as spatial resolution and efficiency. Recent advances in technology (scintillators, photodetectors, and high speed electronics) have renewed interest in TOF PET, which is experiencing a rebirth. However, there is still much to be done, both in instrumentation development and evaluating the true benefits of TOF in modern clinical PET. This paper looks at what has been accomplished and what needs to be done before time-of-flight PET can reach its full potential.

Speaker: Dr William Moses (Lawrence Berkeley National Laboratory)

Slides Tue1500-P2.pdf Tue1500-P2.ppt

16:00 Development of new radiotracers

Speaker: Dr Michael Hofmann (Insel hospital, Bern, Switzerland)

16:30 coffe break

17:00 Imaging with micropattern gas detectors

A new generation of micropattern gas detectors (MPGD) has been developed, mostly to cover the needs of particle physics. The excellent performances in terms of two-dimensional localization and rate capability makes MPGD very attractive for digital imaging of radiation. After a short summary of performances, recent applications in physics, astrophysics, biomedicine will be presented.

Speaker: Fabio Sauli (INFN Trieste and CERN)

Slides Tue1700-P1.pdf Tue1700-P1.ppt

17:30 MICROELECTRONICS TECHNOLOGIES FOR NEW DETECTORS IN MEDICAL IMAGING

Abstract Invited Tutorial Microelectronics technologies play an ever more important role in medical imaging, both in the imaging devices themselves and in the digital capabilities to enhance and analyze the image. This trend will continue in the future, and instrumentation developments in elementary particle physics can serve as a proving ground for new directions in digital imaging detectors. Specific aspects are the use of direct conversion in a semiconductor matrix, innovative 3-dimensional detector construction, very fast signal processing, on-line data pre-processing and massive parallelism at the system level. Besides the general roadmap in CMOS the presentation will discuss recent directions in technology that aim at system integration at the hardware level such as multilayer devices and high density interconnects. With the Medipix development as an example, the ultimate aim of single photon imaging will be discussed. The cost aspects of the semiconductor imager options have to be taken into account in the R&D phase.

Speaker: Erik Heijne (CERN)

18:00 Quantification in emission tomography: challenges, solutions, performance and impact

Objective: Emission tomography (Single Photon Emission Tomography – SPECT – and Positron Emission Tomography – PET) offers a great potential for the quantitative characterization of functional and molecular processes in vivo. Indeed, voxel values in the reconstructed SPECT and PET images can theoretically be translated into a well-understood physical quantity, namely radiotracer concentration, from which all kinds of parameters characterizing molecular processes can be derived using appropriate modeling. Such quantitative interpretation of SPECT and PET images is often referred to as “quantification”. The objective of this tutorial is to examine how far we are in this quest for quantification by presenting the different problems that have to be addressed, the methods that have been developed to solve these problems, and the current performance of these methods. Contents: To successfully achieve quantitative interpretation of SPECT and PET images, the very first step consists in establishing a linear relationship between radiotracer concentration and voxel values. To do so, many phenomena have to be taken into account, among which motion, photon attenuation and photon scatter, non stationary spatial resolution and partial volume effect are the most critical. For each of these phenomena, the resulting biases will be explained and illustrated in different configurations, the most relevant compensation methods will be described and the current performance that can be achieved using these methods will be presented. The influence of the random correction in PET, of the tomographic reconstruction algorithm, and the very way measurements are performed from the images will also be discussed. The respective importance of the different factors affecting the accuracy with which activity concentration can be estimated from the images will be illustrated. Once proper methods have been used to ensure that radiotracer concentration can be accurately deduced from voxel values, a second step in the quantification process is the estimation of physiological parameters from radiotracer concentrations measured in different physiological compartments and possibly over time. Such estimation is based on modeling the physiological system under study. Examples will be shown to demonstrate that relevant modeling is of foremost importance to take full advantage of accurate estimates of radiotracer concentration from the SPECT and PET images and successfully complete the quantification process.

Speaker: Mrs Irène Buvat (U678 Inserm UPMC)

Slides Tue1700-P3.pdf Tue1700-P3.ppt

19:00 → 20:00 Welcome Reception

Wednesday 10 May

08:30 → 12:50 Small animal imaging

08:30 Progress in Small Animal Imaging

The growing field of molecular imaging has accelerated the introduction of novel technologies and the development of new methodologies to investigate biologic processes in living animals that can be used as models of human diseases. Because of their unrivalled sensitivity, radiotracer-based imaging modalities such as Single Photon Emission Computed Tomography (SPECT) and Positron Emission Tomography (PET) enable the noninvasive investigation of biochemical and physiological processes in vivo at the cellular and molecular levels without pharmacological side effects. However, the importance of obtaining the anatomical context alongside with the radioisotopic images is highlighted as new molecular probes capable of targeting receptors and gene expression with high specificity in selected cells and tissues are developed. In this overview, we will briefly review the most recent developments in SPECT and PET imaging, as well as survey the new multi-modality imaging methodologies being explored to provide more complete anatomical, physiological and molecular information concurrently in small animals.

Speaker: Prof. Roger LECOMTE (Université de Sherbrooke)

09:00 Initial Studies Using the RatCAP Conscious Small Animal PET Tomograph

The RatCAP is a small, head mounted PET tomograph that has been designed to image the brain of a conscious rat. It will allow PET imaging studies to be carried out on laboratory rats without the use of anesthesia, which severely suppresses many of the brain functions and neurological activities that one would like to study using PET, and also prevents studying animal behavior while simultaneously acquiring PET data. The RatCAP is a complete, high performance 3D tomograph consisting of a 3.8 cm inside-diameter ring containing 12 block detectors, each of which is comprised of a 4x8 array of 2.2x2.2x5 mm3 LSO crystals read out with a matching APD array, and has a 1.8 cm axial field of view. The APDs are read out using a custom designed ASIC and FPGA time stamp module using a VME readout system. The complete detector weighs less than 200 gram and is suspended by a tether which allows limited freedom of movement of the awake animal. The first working prototype of the RatCAP has been studied using various phantoms and was determined to have a point source resolution of 2.1 mm FWHM (measured in the center of the field of view), a time resolution of 13.9 ns FWHM , and an overall sensitivity of 0.7% at an average energy threshold of 150 keV. Initial brain images of live rats have been obtained and show comparable resolution to similar images obtained with a commercial microPET R4 scanner. Initial studies have also been carried out to study stress levels in rats wearing the RatCAP, and to improve their adaptation and tolerance to the device through training and positive food reinforcement. Results on improved system performance, along with additional brain and animal behavioral studies, will be presented at the conference.

Speaker: Dr Craig Woody (Brookhaven National Lab)

09:15 Evaluation of the performance of the YAP-(S)PET scanner and its application in neuroscience

Objectives: The YAP-(S)PET is the only small animal scanner that combines the PET and SPECT techniques on a single gantry. It is made up of four detector heads: each one is composed of a 4 cm × 4 cm YAP:Ce matrix of 20 × 20 elements, 2 mm × 2 mm × 25 mm each, coupled to a PS-PMT. Due to the relatively large solid angle subtended by each crystal matrix the system is subject to an appreciable pile-up probability. In order to improve the count rate capabilities, a completely new and faster readout electronics has been installed. The new electronics offers a reduced system dead time leading to a maximum acquisition rate ten times higher than the previous one. The new circuitry also includes a proprietary pile-up rejecting technique. In this work we present the performance evaluation in terms of absolute sensitivity, spatial resolution and count rate capability, both in PET and SPECT modalities, of this new version of the scanner which is installed at the CNR-IFC in Pisa (Italy) within the framework of the Center of Excellence AmbiSEN of the University of Pisa. Methods: In PET mode a 22Na point source (1 mm Ø) was positioned at CFOV and moved radially with 0 mm, 5 mm, 10 mm offset. For each position, the radial, transaxial and axial FWHM and FWTM are measured. The exact methodology for measuring the scatter fraction and count rate performance (NEC) in small animal PET scanners is still an open question. The PET performance assessment described here are then based on the preliminary standards proposed by the small animal PET NEMA task force. The scatter fraction and NEC curve were evaluated using a mouse-like phantom filled with a known quantity of activity of a 18F solution. The PET system sensitivity is measured with a linear source placed inside a metal tubes. The measure is repeated five times with increasing wall thickness. The effectiveness of the new pile-up rejection technique has been evaluated by comparing the uniformity of the reconstructed images of an uniform phantom, at various count rates, with and without the pile-up rejection capability. In SPECT mode a glass capillary filled with 99mTc solution was used for the evaluation of both spatial resolution in the transaxial plane and sensitivity. Images of a mini Derenzo phantom, with hot rods down to 1.2 mm diameter, and of the NEMA PET image quality phantom will be presented for both PET and SPECT modalities. Animal imaging: A new model of Parkinson disease has recently been developed on the EnHT mice. To confirm and extend the previous analysis on the EnHT mice as a new model for Parkinson’s disease we evaluated the binding of DatSCAN as indicator of integrity of dopaminergic system in asintomatic EnHT mice. SPECT image was collected 4 hours after tail-vein injection of about 80 MBq of 123I-FP-CIT. Images was reconstructed with EM using a dedicated procedure for 123I.

Speaker: Dr Nicola Belcari (Department of Physics, University of Pisa)

09:30 Status of the optimized ClearPET(tm) Neuro scanner

The ClearPET(tm) Neuro scanner, a dedicated high performance small animal PET scanner at Forschungszentrum Jülich GmbH, has been developed as a device for molecular neuroimaging in rodents and non-human primates under physiological conditions. Innovative features developed within the Crystal Clear Collaboration (CCC) have been implemented in particular with regard to crystal material and electronics which allow achieving both high resolution and high sensitivity. Each of the 80 detector modules consist of a crystal matrix with two crystal materials, LSO and LuYAP:Ce, combined in a phoswich arrangement which are coupled to a multi-channel PMT. The gantry allows rotation of the detector modules as well as tilting by 90 degrees to measure non-human primates in an upright sitting position. The opening diameter of the ring is variable between 130mm and 300mm, the axial detector length is 110mm. The reconstruction of the images has been performed by using the 3D-OSEM algorithm based on the STIR library (Software for Tomografic Image Reconstruction, http://stir.irsl.org <http://stir.irsl.org/> ). A suitable normalization procedure takes account of the unconventional geometry. Already the very first measurements of phantoms and also with rats confirmed the high performance of the scanner and showed its capabilities for in vivo small animal scanning. Based on these preliminary results during the first testing periode the ClearPET(tm) Neuro System has been modified to optimize specific properties like the DOI, the detector positions and the maximum count rate. For the new assembly the crystal matrices have been repackaged in firmer and more compact blocks with BaSO4 as reflector material between crystals. The axial distance between the detector blocks within the detector cassettes has been reduced and the shift between every other cassette has been changed. The new positioning of the detectors was implemented to achieve a higher packaging density and a distance between the detector blocks which is an integer multiple of the slice width. To improve the count rate performance further a new file server is implemented which will allow a maximum count rate of 9 Mcps single events. We present the current status of the ClearPET Neuro scanner and elaborate on the effects of the recently implemented modifications.

Speaker: Dr Karl Ziemons (Forschungszentrum Juelich)

09:45 SiliPET: An ultra high resolution Design of a Small Animal PET scanner based on Stacks of Double Sided Silicon Strip Detector

Positron Emission Tomography (PET) of small animals is a good and promising in-vivo and non-invasive technique which has helped the recent emergence of molecular imaging. The fast development of this research fields needs new technological devices to reach better and better performances in term of spatial resolution and sensitivity. To satisfy these growing requests, small animal PET scanners, based on pixellated matrices of scintillator crystals coupled to position sensitive photomultipliers, have been pushed to use always smaller pixel sizes. This strategy cannot go much further because it is approaching both physical and technical limits: inter-crystal scatter of incident photons; energetic fluorescence X-rays produced by high Z material used in scintillators for high efficiency together with a high photoelectric fraction; difficulty in distinguishing the always smaller pixel elements in a crystal matrix using position sensitive phototubes; parallax error due to the depth of interaction in thick crystals. To improve these imaging devices it is mandatory to change the approach and move towards new solutions. We have studied with Monte Carlo simulations, using the EGSnrc code, a new scanner for small-animal positron emission tomography (PET) based on stacks of double sided semiconductor detectors. In small animal PET imaging (mice) Compton scattering within the animal itself is not an issue and therefore registration of the energy information is not necessary. Low Z materials can therefore be studied for designing a high performance scanner. Each stack is composed of planar detectors with dimension 60 x 60 x 1 mm^3 and 128 orthogonal strips on both sides to read the two coordinates of the interaction, the third coordinate being the detector number in the stack. Coincidence events are recorded only if two plane detectors in two different stacks registered an energy deposition over threshold. In this way we overcome three of the crucial scanner limits: depth of interaction measurement; inter-detector multiple interactions by requiring a single hit; reduced non colinearity effects. This configuration results in a large solid angle coverage for high sensitivity. This is a promising new approach for small animal PET imaging. The sinogram profile for a pointlike source of 18F in a 3 cm diameter water phantom is 0.52 mm FWHM. The simulated sensitivity at the center of the FOV is 5.1% with 4 stacks each 4 cm thick placed in a box-like configuration. Preliminary results of a proof of principle measurement were done with the MEGA advanced Compton imager confirming our Monte Carlo results. The results of simulations and of measurements with MEGA imager will be presented, hopefully with some first tests with our prototype detector.

Speaker: Dr Giovanni Di Domenico (Universita' di Ferrara & INFN-Sezione di Ferrara - Italy)

10:00 Scanner Calibration of a Small Animal PET Camera Based on Continuous LSO crystals and a Flat Panel PSPMTs.

We have constructed a small animal PET with four identical detector modules, each consisting of a continuous LSO crystal attached to a Flat Panel PSPMT. The dimensions of the continuous crystal are 50 × 50 mm2 and 10 mm thickness. The modules are separated 11 cm between each other in the scanner. In this paper we discuss the method used for the calibration of the camera for this spetial system with continuous detectors and the first results obtained with simple phantoms and the Derenzo phantom. We also present the preliminary values for the main performance parameters such as spatial and time resolution, energy resolution, and random and scatter fractions.

Speaker: Dr José Benlloch (IFIC, Valencia)

10:15 Expected performances of the AMISSA PET component.

A new Multi Modality Imaging System for Small Animal (AMISSA) is under development, at the Hubert-Curien Multidisciplinary Institute (IPHC, France). Within this framework, a new PET design is under study[1]. This system is aimed to achieve a spatial resolution of 1 µl in the entire field of view (FOV), along with a high detection efficiency, close to the geometrical efficiency in the center of the FOV. To fulfill this requirement, the scanner is designed to give access to the "depth of interaction" information. It is organized into several modules of LYSO crystals. Each module is oriented along the scanner axis, and it is read, at both ends, by a multianode PMT. This layout gives a direct access to the depth of interaction from the crystal address, while the axial position is measured from the light sharing between both crystal ends. The modules are layered into several concentric rings to reach the required detection efficiency. Results of the Monte-Carlo simulations used to define the design of the detector are presented in this paper. Investigations are made on the geometrical layout and on the acquisition system. The expected spatial resolution and counting rate are evaluated and used as criteria for this design process. One major interest of measuring the depth of interaction is to avoid the degradation of the spatial resolution for the points source located at the edge of the field of view. The evolution of the spatial resolution according to the position of the source will then be evaluated with an usual reconstruction algorithm. On the other hand, the impact of the high counting rate, expected from the high detection efficiency will be explored. In particular, the consequences on the timing performances required for the crystal response, the electronic readout and the acquisition elements will be evaluated. The amount of disk space and the minimal data transfer bandwidth necessary to record all the single events will be evaluated.

Speaker: Dr David Guez (IPHC)

10:30 Coffe break

11:00 FDG-PET scan shows increased cerebral blood flow in rat after sublingual glycine application

Isotope studies provide valuable data about an organ’s function in vivo. Thanks to positron emission tomography (PET) using the radiolabeled natural metabolites, such as [18F]-2-fluoro-deoxy-d-glucose (FDG), biological and physiological meaning of nuclear medicine scans has been considerably increased. Moreover, the higher spatial resolution of PET scans allows the direct assessment of metabolism even in small animals. Therefore it is of interest to elucidate the possibilities of the technique in a study of some natural metabolites like glycine influencing the blood microcirculation. As a medicine glycine was recently shown to have a positive therapeutical effect in the treatment of patients with ischemic stroke and some other neurological disorders based on vascular disturbances. By previous direct biomicroscopic investigations of pial microvessels in laboratory rats an expressed vasodilatory effect of topically applied glycine was proved. The aim of this study was to evaluate an influence of glycine on cerebral blood flow (CBF) in rat using FDG-PET scan. A baseline study was started immediately after IV injection of 19 MBq of FDG in a rat anesthetized with ketamine and xylazine. The PET images were acquired twice, one by one during 20 min. Two hours later, after sublingual application of glycine (200 mg) and the second FDG injection, the pair of PET scan was performed during 20 min as well. Finally, 4 days after the first studies, we repeated the PET scans in the same conditions after sublingual application of glycine (200 mg). The quantitative analysis of FDG volume concentration (Bq/ml) in the rat brain was performed based on maximum, minimum, mean and standard deviation values comparison (e.soft software, Siemens, Germany). The quantitative analysis demonstrated that, in two studies after glycine administration, FDG volume concentration in the brain increased at least 1.5 times in comparison with the baseline data. Moreover, the peak of FDG volume concentration in the brain was coming in more rapidly. In fact, after glycine dosing, this peak was observed in the scans performed immediately after FDG injection and not in the second scan seen in the baseline study. These results confirm enhancing effect of glycine on the rat CBF possibly because of its vasodilatory effect on brain microvessels. Therefore, FDG-PET technique contributes to better understanding of glycine pharmacokinetics and its positive influence on brain blood supply. Further studies involving microPET technique and radiolabeled glycine are needed for following up the fate of this natural metabolite in vivo condition.

Speaker: Dr Oleg Blagosklonov (University of Franche-Comte and Jean Minjoz University Hospital, Besançon, France)

11:15 Monte Carlo simulation of small rodents imaging applications using the microPET FOCUS

Positron Emission Tomography is an essential modality for small animal molecular imaging. Growing requirements in high animal throughput and precision imply the optimization of acquisition parameters and protocols. Monte Carlo simulations are essential tools for assisting these developments, improving image quantification and data analysis. In this work we use the GATE platform [1] based on the Geant4 toolkit package and dedicated to nuclear medicine imaging. This platform is well suited for modelling the microPET FOCUS system dedicated to small animal PET imaging and for implementing realistic phantoms for small animal imaging. In this work, we validate the microPET FOCUS modelling with GATE, for spatial resolution, counting rate performances, image contrast recovery and quantitative analysis. Results of realistic D2-receptor studies (using [C-11]Raclopride) and glucose metabolism studies of the rat brain using realistic Monte Carlo simulations of dynamic studies are presented. These simulations include realistic injected doses to the animal and realistic time framing. We show the relative contribution of positron range and γ accolinearity on the quantitative analysis of rat brain functions in various regions. [1] Jan S et al “GATE: a simulation toolkit for PET and SPECT”, Phys. Med. Biol. 49 (2004) 4543-4561.

Speaker: Dr Sébastien jan (CEA/DSV/DRM/SHFJ - Orsay France)

11:30 New trends in Gamma-Ray Imaging with CdZnTe/CdTe

Increasing interest in room temperature solid state detectors like CdTe/CdZnTe for gamma-ray imaging has marked the last ten years with progress in crystal growth, device technology, integrated electronics and signal processing. The purpose of this paper is to present recent technological advances and potential applications in gamma-ray imaging. A new method to grow very large dimension CZT up to 300 mm in diameter is presented. Metal-Semiconductor contact process control allows design of very small electrodes (up to 100µm pitch) with good long-term stability. The development of low-noise front-end integrated electronics and new interconnects technologies have made “pixel detectors” feasible. Significant effort has been focused on reaching high-energy resolution as well as high detection efficiency. The combination of signal processing such as bi-parametric (BP) approaches and electrode design presents an interest in charge loss compensation, tailing reduction due to geometric effect and depth of interaction information. Regarding improvement in CZT performance, molecular imaging (SPECT and PET) is the most promising field of gamma-ray application and several developments of CdTe/CZT based solid state gamma camera have emerged with the expectation that such new technology would outperform conventional scintillator detectors like NaI(Tl) or LSO crystals. In this framework, CEA-Leti is being working on CZT technology and system approach for gamma-ray imaging that may still improve image quality either for low or high energy applications. An energy resolution of about 1.6 % at 122 keV with a detection efficiency of 75% has been obtained with pixelated 10x10x5 mm3 CZT detectors that allows both multi- isotopes identification and scatter event rejection. Such new CZT performance (high energy resolution, high intrinsic spatial resolution, depth of Interaction information and high count rate capabilities) may open up both new vistas for innovative system architecture of detector / collimator to overcome the sensitivity- spatial resolution trade off of the conventional Anger gamma camera and new image processing with multi-energy approach. New system architecture and image processing that take the benefit of CZT performance will lead to a technical breakthrough with the standard procedure used today in nuclear medicine either for clinical (SPECT) or small animal imaging (µSPECT). To address the spatial resolution homogeneity issue associated with small animal µPET systems, we have also jointly developed a specific three-dimensional CZT detector with orthogonal strip geometry for a transverse irradiation and a preamplifier stage: an encouraging CZT-BaF2 coincidence time of 2.6ns FWHM at 511 keV has been achieved and a good homogeneity of the spatial resolution across the Field Of View (<1 mm FWTM up to 44 mm off the FOV center) is expected. This new 3D CZT detector geometry may open up new vistas for innovative architecture in µPET system, particularly regarding the sensitivity improvement.

Speaker: Dr LOICK VERGER (CEA-LETI)

11:45 MediSPECT: Single Photon Emission Computed Tomography System for Small Field of View Small Animal Imaging Based on a CdTe Hybrid Pixel Detector

We describe MediSPECT a newly developed system for SPECT studies on small animals with a small field of view (FOV) and high spatial resolution. The detection imaging system is based on a CdTe pixel detector (a 256x256 matrix of 55 micrometers square pixels) operating in single photon counting for direct detection of X-rays and gamma- rays with low and medium energy (e.g. I-125, 27-35 keV, Tc-99m, 140 keV). The hybrid detector is obtained by bump-bonding, pixel by pixel, the semiconductor detector to the Medipix2 read-out chip developed by the Medipix2 collaboration. The detection imaging system can be coupled to interchangeable tungsten collimators with very high resolution: a parallel hole collimator with circular holes of 100 micron and 70 micron septa; knife-edge pinhole collimators with a diameter of 0.35 mm, 0.78 mm, 1 mm and 2 mm (90° aperture); coded aperture masks with 70 micron and 80 micron holes for low energy imaging. At present, the useful FOV of the MediSPECT scanner ranges from 6.3 mm with the coded aperture mask (system spatial resolution 120 micron @ 27- 35 keV) to 31 mm with 0.35 mm pinhole (system spatial resolution 1 mm @ 27-35 keV and 2.3 mm @ 140 keV). These features make MediSPECT an interesting system for in vivo imaging of small organs or tissue structures in mouse, e.g., brain, thyroid, heart or tumor. A rotating gantry hosting the detector and a precise system for collimator alignment allows for 360° rotation around the horizontal animal bed with a radius of rotation ranging from 20 mm to 100 mm. At present we have tested the system only with one detector head, but it is also possible to mount a second detector/collimator unit for increasing sensitivity. MediSPECT is controlled by a dedicated software, to control the hybrid detector functioning, the tomographic head alignment and rotation, the animal bed movement and the image acquisition as planar projection for off-line SPECT reconstruction. We show preliminary images obtained with point-like sources and sample phantoms (Cd- 109 22 keV, I-125 27-35 keV, Tc-99m 140 keV) acquired with MediSPECT both in planar and tomographic configuration. A detailed description of the 3D image reconstruction algorithm is reported in a companion work presented at this conference.

Speaker: Dr Maria Cristina Montesi (Dipartimento di Scienze Fisiche, Università di Napoli Federico II and INFN sez. Napoli, Napoli, Italy)

12:00 The First Four-Headed NanoSPECT: A High-Sensitivity Multi-Pinhole SPECT System for Imaging Small-Animals with Submillimeter (Nanoliter) Spatial Resolution

Multi-Pinhole SPECT has become a proven modality in small-animal molecular imaging. Although the spatial-resolution capabilities of SPECT are greater than those of PET, the latter is generally considered the gold-standard nuclear imaging modality due to high-sensitivities. In this work, we present a high-throughput SPECT system that achieves submillimeter reconstructed resolutions while simultaneously approaching the sensitivity of PET. This increase in sensitivity combined with existing advantages of SPECT, e.g. tracer chemistry, cost and dual isotope capabilities, improves the standing of SPECT as a molecular imager. This camera, the NanoSPECT, consists of four detectors (215x230mm2 NaI, 33 PMTs, 2.1mm intrinsic resolution at 140keV) mounted on a high-precision gantry. Each detector is outfitted with an interchangeable 9-pinhole aperture for a total of 36 pinholes surrounding the field of view (FOV). Pinhole diameter and FOV are chosen in accordance with the prescribed application, e.g., mouse or rat imaging. The axial FOV is extended using helical scanning (user-selectable range from 20 to 290mm). Additionally, helical orbits provide an increase in the angular sampling. All told, this increase in sensitivity and sampling greatly improves image quality both for detection and estimation (quantification) as compared to standard SPECT acquisition techniques. We will present a detailed description of the NanoSPECT along with numerous phantom studies and small-animal scans performed with an array of Tc-99m, I-123 and In-111 tracers. The results will address resolution, sensitivity, imaging times, injected dose and quantification results as well as multi-isotope and dynamic SPECT capabilities.

Speaker: Dr Christian Lackas (Research Center Juelich)

12:15 Performance evaluation of a mouse-sized camera for dynamic-studies in small animals

A mouse sized camera has been built in terms of a collaboration between the presenting insitutions. The system will be used for the performance of dynamic studies in small animals, in order to evaluate novel radiopharmaceuticals. The active area of the detector is approximately 48 mm x 96 mm allowing depiction of the entire mouse in a single view. The system is based on two flat-panel Hamamatsu H8500 position sensitive photomultiplier tubes (PSPMT), a pixellated NaI(Tl) scintillator and a copper-beryllium (CuBe) parallel-hole collimator. In this work the evaluation results of the system are presented, using phantoms and small animals injected with conventional radiophrmaceuticals. Average resolution was ~1.6mm on the collimator surface and increased to ~4.1mm in 12cm distance from the detector. The energy resolution was measured and found ~15.6% for Tc99m. Capillary imaging demonstrated system's high resolution and sensitivity in activity variations. Initial dynamic studies have been carried out in small animals injected with Tc99m-DTPA and Tc99m- MDP. Successive images were acquired. Counts in ROIs were plotted as a function of time. The results show systems ability to perform kinetic imaging in small animals. Currently dynamic studies are carried out in mice and rats injected with Tc99m- Bombesin derivatives and results will be presented as well. As a next step system's performance will be tested using two pinhole collimators and initial tomographic studies will be carried out using a rotating table and phantoms. Future upgrading of the system to an integrated SPECT/CT is planned.

Speaker: Dr George Loudos (Biomedical Simulation and Imaging Application Laboratory, School of Electrical and Computer Engineering, National Technical University of Athens)

13:00 → 14:00 Technical Presentation Siemens Medical Solutions : MI on MR/PET Solutions

Speaker: Dr Olivier De Dreuille (Siemens Medical System)

14:00 → 14:30 Presentation pole Photonique

14:00 → 15:00 Poster session : detection modules and electronics

14:00 A Front-End Readout Mixed Chip for High Efficiency Small Animal PET imaging.

Today, the main challenge of Positron Emission Tomography (PET) systems dedicated to small animal imaging is to obtain a high efficiency event data collection with a good sensibility and a high accurate localization. If the focus is only on the PET characteristics, an accurate spatial resolution depends as well on the design of detector as its electronics readout system. In this paper, we present a new design of such readout system with full custom sub-micrometer CMOS implementation. The chip consists of two main blocks in which the energy information and temporal nanosecond resolution data can be obtained. In our AMISSA PET device design, a matrix of LYSO crystals has to be read at each end by a 64 channels multi-anode photomultiplier tube. Then a specific readout electronic has been developed at IPHC. The architecture of this readout for the energy information detection is composed of a low noise preamplifier, a CR-RC shaper and an analog memory. In order to obtain the required dynamic range from 0.3 to 650 photoelectrons with good linearity, a current mode approach has been chosen for preamplifier. To detect only the useful signal with a temporal resolution of one nanosecond and to eliminate the interference due to the afterglow generated by the crystal, a double in time comparing system with a low threshold (0.3 photoelectron) has been implemented. It gives the time reference of arrival signal coming from the detector. In order to obtain the time coincidence with a temporal resolution of one nanosecond, a Time-to-Digital Converter (TDC) based on a Delay- Locked-Loop (DLL) has been designed. The chip is fabricated with AMS 0.35 m process. The ASIC architecture and some simulation results will be presented in the paper.

Speaker: Mr Nicolas OLLIVIER-HENRY (Institut Pluridisciplinaire Hubert-Curien, UMR 7178 CNRS/ULP)

14:00 A functionally reconfigurable detector head for molecular imaging with radionuclides

Molecular imaging by radionuclides is a very powerful technique to study mouse models and to understand fundamental genetic and biochemical processes. The development of compact, high-resolution gamma cameras, based on both single photon (SPECT, E=140 keV) and positron emission (PET, E=511keV) techniques calls for dedicated compact detectors characterised by both high spatial resolution and high sensitivity. Several prototypes have been successfully developed [1-4 ] but studies aiming at identifying optimal solutions are still under way. In the present work, we report a new detector head based on a two-dimensional array of PIN-diode photo detectors. PIN diodes, besides their intrinsic insensitivity to magnetic fields, are technologically simple devices, do not need of high bias voltages, have excellent linearity and quantum efficiency and have energy resolution that, at room temperature and for long integration times, is comparable with that of drift detectors for 140 keV gamma-photons [5]. The detector is made of a two-dimensional array of 2mm x 2mm 64 PIN diodes integrated on high-resistivity silicon, coupled with a scintillator on the back side, and bonded to mixed-mode CMOS read-out on the front side. The whole system has a high modularity so as to be adapted to larger array dimensions with 1mm x 1mm pixel sizes. It includes an FPGA based PC board for routing the electronic signals and for interfacing the detector with a supervisor PC. The front-end electronics exhibits fairly good performance in terms of equivalent input noise charge (ENC < 100 e-rms) and uses a novel architecture for the baseline holder circuitry [6-8]. Digital transmission is realised through a custom bidirectional serial protocol, with multiple parity checks. Data can be acquired and transmitted to the main board in different ways, since multiple chip acquisition modes are supported. This allows changing the modalities and the times to form a scintimammography image. In SPARSE_READOUT and TEST modes, event positioning and amplitudes are acquired and transmitted out of the module to the main board. These modes are particularly useful for detector calibration and testing, although they do not provide maximum performance since the transmission requires about 60 clock periods. However, in SPARSE_READOUT mode the amount of data is usually much smaller since the amplitude word is only transmitted whether the corresponding pixel appears to be excited. WTA (Winner Take All) and FTA (First Take All) modes help minimising data transfers while gathering important information about those pixels on which the maximum photon incidence is found. For low event rates, WTA [9] provides better accuracy in the measure of the maximum amplitude level among pixels. When speed is of some concern, FTA acquisition mode is preferred since it takes about 20 clock cycles to send complete information concerning the excited pixel. Moreover, speed up of the transmission is also possible by reducing the amount of information sent: for this reason, 8 distinct WTA and FTA modes are available. Further operating modes have been implemented to allow the acquisition of data even if the pixels are not excited, providing a useful method for detector calibration and testing. The front-end chip has been implemented in AMS 0.35um CMOS technology, while the readout and control logic has been implemented, using an Altera Cyclone FPGA [10]. The whole system has been tested on an interface board and it has proven to work successfully even at the maximum supported clock speed of 12MHz. References [1] W.-S. Choon, G. J. Gruber, W. W. Moses et al. “A Compact 16-Module Camera Using 64-Pixel CsI(Tl)/Si P-I-N Photodiode Imaging Modules”, IEEE Trans. Nucl. Sci., vol. 49, no. 5, pag. 2228-2235, October 2002. [2] G. J. Gruber, W. S. Choong, W. W. Moses, S. E. Derenzo, S. E. Holland, M. Pedrali-Noy, B. Krieger, E. Mandelli, G. Meddeler, and N. W. Wang, "A compact 64- pixel Csl(Tl)/Si PIN photodiode imaging module with IC readout", IEEE Trans. Nucl. Sci., vol. 49, pag. 147-152, 2002. [3] B. E. Patt, J.S. Iwanczyk, C. Rossington Tull, N.W. Wang, M.P. Tornai, E.J. Hoffman, “High Resolution CsI(Tl)/Si-PIN Detector Development For Breast Imaging”, IEEE Trans. Nucl. Sci., vol. 45, no. 4, pagg. 2126-2131, August 1998. [4] L. Stebel, M. Tommasi, S. Carrato, G. Cautero, M. Petasecca, G. Pignatel, C. Marzocca, A. Tauro, A. Dragone, F. Corsi, G. Dalla Betta, A. Fazzi, N. Zorzi, “ Development of a prototype detector for use in scintimammography imaging”. First Int. Workshop on Advances in Sensors and Inerfaces, IWASI 2005, Bari, 19-20 April 2005. [5] Fiorini C., "Gamma detectors for spectroscopy and imaging based on scintillators coupled to silicon photodetectors", Proceedings Spie Vol. 4141 (2000) 97-110. [6] G. De Geronimo, P. O’Connor and J. Grosholz, “A CMOS baseline holder (BLH) for readout ASICS”, IEEE Trans. Nucl. Sci., vol. 47, n. 3, pag. 818-822, June 2000. [7] M.W. Kruiskamp, O.M.W. Leenaerts, “A CMOS Peak Detector Sample and Hold Circuit”, IEEE Trans. Nucl. Sci., vol. 41, n. 1, pag. 295-298, February 1994. [8] J. H. Huijsing, R. Hogervorst, K. J. de Langen, “Low-power low-voltage VLSI operational amplifier cells”, IEEE Trans. on Circuits and Systems I, vol. 42 , n. 11, pag. 841-852, November 1995. [9] W. W. Moses, E. Beuville, M. H. Ho, “A Winner-Take-All IC for determining the crystal of interaction in PET detectors”, IEEE Trans. Nucl. Sci., vol. 43, pag. 1615- 1618, 1996. [10] http://www.altera.com

Speakers: Prof. Francesco Corsi (DEE - Politecnico di Bari), Prof. Giorgio Pignatel (Univerità di Perugia)

14:00 A systematic study of the performance of the CsI:Tl single-crystal scintillator under x-ray excitation

The light emission performance of the x-ray excited CsI: Tl single crystal scintillator was investigated as a function of x-ray tube voltage and crystal thickness. Five CsI: Tl single crystal layers (CRYOS Ltd., Ukraine) with thickness from 1 to 7 mm were irradiated employing two x-ray tube voltage ranges: (i) The 22 - 45 kV (molybdenum anode-molybdenum filter (Mo/Mo)) range, employed in mammographic imaging and (ii) The 40-140 kV (tungsten anode-aluminum filter) tube voltage range, used in general x-ray projection and tomographic imaging. The x-ray luminescence efficiency (light emission spectrum area over incident x-ray fluence) of the crystals was determined by performing light emission spectrum and x-ray exposure measurements. In addition the intrinsic quantum gain (photons created per absorbed x-ray) and the spectral compatibility to various optical detectors were estimated from these measurements. The luminescence efficiency was found to be a non-linear function of crystal thickness and of x-ray tube voltage. Peak efficiency (30.7 ) was observed for the 5 mm thick crystal at 140 kV. A secondary efficiency peak was observed at 42 (Mo anode) probably due to the effect of the K-photoelectric absorption edge (at 33 and 35 for Cs and I respectively. For the thicker (6 mm, 7 mm) crystals the efficiency was found to decrease due to light attenuation effects within the scintillator mass.

Speaker: Prof. Ioannis Kandarakis (Department of medical instrumentation-Technological Educational Institution of Athens, 12210 Egaleo, Greece)

14:00 A Waveform Sampling Front-End (WSFE) ASIC for Readout of GSO/APD with DOI Information

Individual readout of detectors in Positron Emission Tomography (PET) scanners give a better spatial resolution and counting rate over light sharing or charge division schemes. However, this will greatly increase the number of readout channels and is difficult to implement with bulky conventional front-end electronics especially with small detector systems like animal PET. To achieve individual readout of large numbers of channels, the use of compact front-end electronics, especially when space is restricted, are required. Application Specific Integrated Circuit (ASIC), being highly integrated, power efficient and reliable, is a promising alternative to conventional front-end electronics while allowing simple integration of analog and digital components. A high resolution avalanche photodiode (APD) based animal PET with depth-of- interaction (DOI) information has been proposed. Each detector module consists of two or more crystals with different decay times, for example GSO doped with different Ce concentrations, that are stacked as in a phoswich detector and each crystal of the lowest layer is coupled to each pixel of a multi-element APD for individual readout. An improved 9-channel (including 1 test channel) Waveform Sampling Front-End (WSFE) ASIC based on a previous design has been fabricated using Rohm 0.35u CMOS technology for readout of signals from each pixel at an early stage. Each channel of the ASIC comprises of a preamplifier, a variable gain amplifier (VGA), a fast ADC and digital encoder to digitize the incoming signal which can then be used to obtain DOI information by pulse shape discrimination (PSD). Each component was analyzed using the test channel. The preamplifier is optimized for low capacitance APD and is based on the telescopic-cascode topology. It has a gain of 1.4/pF and the linearity is less than 0.5% over -0.35 pC to 1.1 pC. The rise time is 20ns and the minimum Equivalent Noise Charge (ENC) is 480 e- + 20 e-/pF rms at a shaping time of 0.25 us. A single-stage VGA instead of the two- stage VGA as in the previous chip was adopted to eliminate offset errors as observed before. The gain can be varied from about 5 to 16 by two external digital inputs. The ADC is identical to that of the previous chip. It is a folding ADC which works at least up a rate of 100 Msamples/s. The DNL and INL were calculated to be 1.1 LSB and 1.4 LSB respectively. To test for the whole channel, two pulses with rise times of 50 ns and 100 ns were fed into the preamplifier. It was shown that the rise times of the two digitized and amplified signals could be distinguished from each other. Finally, the power consumption of the whole chip was measured to be about 1 W when working at 100 MHz. GSO crystals with Ce dopant concentration of 0.5, 1.0 and 1.5 mol% will be coupled to a Hamamatsu APD to examine the DOI capabilities using this WSFE chip. The digitized signals from the WSFE chip can be fed into a FPGA to obtain DOI information by comparing the rise times.

Speaker: Mr Jung Yeol Yeom (University of Tokyo)

14:00 Characterisation of Geiger mode avalanche photodiodes for medical imaging applications.

Recently developed multipixel Geiger-mode avalanche photodiodes (GAPDs) are very promising candidates as detectors of the light from scintillating crystals or fibers in medical imaging (e.g. positron emission tomography) as well as in high energy physic and astrophysical applications. They are specially well suited for morpho-functional imaging (multimodality PET/CT, SPECT/CT, PET/MRI, SPECT/MRI ...). GAPDs have many advantages over conventional photosensors such as photomultiplier tubes because of their compact size, low power consumption, high quantum efficiency and non-sensitivity to magnetic fields. Compared to avalanche photodiodes and PIN diodes they are advantageous because of their high gain and reduced sensitivity to pick up noise. We present the results of measurements of the basic GAPD characteristics: photon detection efficiency, gain, inter-cell crosstalk, dynamic range, timing resolution, recovery time, dark count rate and dark current.

Speaker: Mr Ilia Britvitch (ETH, Zurich, Switzerland)

14:00 Characteristics of Optical Imaging Capillary Plate Gas Detector with Fibre Optics Plate

A capillary plate (CP) gas detector with fibre optics plate (FOP) filled with argon gas mixture and pure xenon gas has been successfully operated as a gas scintillation proportional counter. Gas gains of up to 104 can be achieved with these gas mixtures. Scintillation light simultaneously emitted during the development of electron avalanches can be observed through the FOP using a photomultiplier tube (PMT) and an optical imaging system which consists of a CP gas detector and a cooled CCD camera coupled to lens optics. Using this optical imaging system, clearer images of X-rays have been obtained with the above gas mixtures. Successful operation with these gas mixtures has allowed us to realize a novel imaging device with a CP for medical imaging, cellular function analysis, and X-ray imaging.

Speaker: Dr Fuyuki Tokanai (Yamagata University)

14:00 Crystal growth and potential utilisation of single crystal fibers in medical imaging devices

Single crystal fibers have become a research attraction in terms of both their fundamental and technological originality for various applications, such as lasers, non linear optics or scintillators. We will expose the growth of scintillating fibers using the micro pulling down technology. This technology allows the growth of single crystal fibers with diameter ranging from 0.5 to 3 mm and with length up to 1 meter. These shape characteristics are difficult to carry out with traditional machining of bulk crystals. Fibers of standard inorganic scintillators crystals have already been grown with this pulling method. We present here the progress of the growth of YAG:Ce, BGO and LYSO:Ce fibers. Our crystals are analyzed in terms of crystal structure, crystallinity and dopant uniformity. Optical and scintillation properties are studied as well and will be presented and compared with that of bulk single crystals. Light propagation through the fibers will also be discussed, since it is a pertinent parameter in fiber based scintillation devices. Last, we will try to show out how such materials can find application in medical imaging devices.

Speaker: Mr benoit hautefeuille (Laboratoire de Physico-Chimie des Matériaux Luminescents (LPCML) UMR 5620, Université Claude Bernard Lyon 1, Domaine scientifique de la Doua, 69 622 Villeurbanne, France)

14:00 Design of a fast, efficient single photon ring for cardiac studies based on LaBr3:Ce module

Radionuclide imaging is a common non-invasive technique used in the evaluation of cardiac function and disease. Currently, planar and SPECT imaging are used in nuclear cardiology to perform gated equilibrium blood pool imaging, myocardial perfusion imaging and first-pass imaging. More than one-third of all nuclear imaging procedures are cardiac imaging and most of these are myocardial perfusion SPECT studies. Therefore, a SPECT system optimized for a dedicated to cardiac imaging could have a significant impact on a large patient population. The standard clinical SPECT imaging using a single-head gamma camera is obviously far from optimal. Several commercial companies have introduced L-shaped camera systems specifically for cardiac SPECT imaging. Although these systems double the coverage of a single-head system and cut the acquisition time for a 180° acquisition in half, there is still a large gap in detection coverage. Among the current systems, the 3-head configuration provides the most complete detector coverage and therefore, the current 3-head configuration with standard Anger cameras of 40 cm FoV is not exactly optimized because it still allows a portion of emitted photons to escape undetected. Intuitively, a cylindrical 2pi geometry should be the optimal configuration to catch the maximum number of photons emitted from a body section. Over the last 20 years several groups have shown interest and proposed designs using cylindrical geometries for body and cardiac SPECT. However, the concept of using cylindrical geometry for cardiac SPECT imaging is not straightforward and needs to be examined carefully. Very recently INFN have proposed a new scintillation crystal for SPECT: LaBR3:Ce. Its superior energy resolution (6.5% at 140 keV) and a short scintillation decay time (16 ns), can allow high counting rate (more than three times higher than NaI(Tl)) and high scatter rejection. It seems an election crystal for cardiac SPECT where very high counting rate are required and a significant fraction (30-40%) of photons detected have been scattered. Preliminary results on continuous LaBr3:Ce have shown 1 mm intrinsic spatial resolution for a crystal thickness corresponding to 80% efficiency at 140 keV photon energy. Measurements on thicker crystals (1 cm) are going to verify spatial resolution values better than 2 mm with a further increasing of detection efficiency of the system. To this aim INFN is promoting a new design of a cylindrical geometry for cardiac SPECT based on LaBr3:Ce detection module. Each module will have one continuous LaBr3:Ce crystal with dimensions of 10cm x 20cm. The optimal crystal thickness (between 5 and 10 mm) is under investigation. Scintillation light from the crystals will be collected by eight 2-in square Position Sensitive Photomultiplier Tubes Hamamatsu H8500 (PSPMTs) optically coupled to each module via a glass window light guide. These modules are designed to be components of a modular cylindrical single photon emission computed tomography system suitable for cardiac imaging. All the 10 modules can be moved along the radial direction and the ring itself, of about 70 cm diameter, can rotate to obtain the required number of views as a function of chosen FoV. Modules can work individually as 10 independent modules with 20cm x 10cm FoV each, viewing the heart centered in the ring, or they can be coupled to form 5 detectors with a slant collimator, with 40cm x 10cm FoV each. In addition, the ring geometry can vary its size or it can form a L shaped camera to be positioned close to the heart to increase system spatial resolution as required in the cardiac gene therapy imaging. Therefore, high resolution collimation is required to compensate for the long target distance at the expense of geometric efficiency. The expected maximum sensitivity of the system is almost 2.7 cps/kBq (100 cps/microCi)with a maximum rate of 2 MHz and a spatial resolution of 2.5 cm. The best spatial resolution would be around 4 mm. Ultimately, the smallest ring geometry consists of 5 modules (20cm x 10cm FoV each) defining a minimum ring diameter of 28 cm. It can be an unique system for paediatric cardiac SPECT.

Speaker: Prof. Roberto Pani (INFN)

14:00 Development Of a New Photodetector Readout Technique

In the framework of the European FP6’s BioCare project, we develop a novel photodetector readout technique to increase sensitivity and timing precision for molecular imaging in PET and CT. Within the Project’s work packages, the CERN- BioCare group focuses on the development of a PET detection head suitable to process data from both PET and CT operation in one unit the latter at a rate of 200 Mcounts cm-2s-1. The detector readout module consists of a LSO matrix coupled to an APD array. The signal is processed by readout electronics recently developed for a Time-Of-Flight detector in the ALICE experiment at the LHC of CERN. The core of this readout electronics is built on a low noise and fast preamplifier that is interfaced to an ultra-fast amplifier-discriminator. We show here the functioning of the individual system components and further discuss the global performance of the entire readout channel, before to conclude on the potential of our proposed technique for both PET and CT applications.

Speaker: Francois Powolny (CERN - 1211 Geneva 23, Switzerland)

14:00 Development of an instrument for time activity curve measurements during Positron Emission Tomography imaging of rodents

Molecular imaging using Positron Emission Tomography (PET) in small rodents commonly requires the knowledge of the input function of the tracer (development of new drugs or new tracers …). In this paper, we report the status and the performances of a prototype of a counting system under development at DAPNIA(1) in collaboration with SHFJ(2). The detection device consists on six silicon diodes of 0.3 mm thickness used to detect the positrons emitted by the radiotracer contained in the arterial blood flowing in a thin wall catheter. Such diodes are poorly efficient for the detection of 511 keV gammas escaping from the rodent and thus require a thin lead shielding. The detectors, the front-end electronics (for signal preamplification, shaping, and discrimination), and the acquisition circuits are mounted on a single card. The device is connected directly to a portable computer via an USB port. Such a design provides a compact, rugged and portable device for working close to a small animal PET camera. Preliminary experimental measurements were performed using 18F solutions with activity concentration close to that encountered during typical animal PET studies. The results show that the detector presents a good detection homogeneity, a good linearity and an efficiency of nearly 40 % close to that predicted using Monte Carlo simulations. This work was partly supported by a grant from IPA 2005-program. 1) Département d’Astrophysique, physique des Particules, physique Nucléaire et Instrumentation Associée 2) Service Hospitalier Frederic Joliot

Speaker: Mr Jean-Marc Reymond (CEA/DSM/DAPNIA)

14:00 Development of scintillation detectors based on avalanche microchannel photodiodes

Avalanche Microchannel PhotoDiodes (AMPDs) are solid state photodetectors with high internal gain and density of independent channels up to 10^4/mm^2. They are potential substitutes for photomultiplier tubes in a wide variety of applications in nuclear physics and nuclear medicine. The use of AMPDs is most promising when fine segmentation of the detectors and their operation in high magnetic fields is required. Here we present our ongoing developments of AMPDs and scintillation detectors based on them focusing on position sensitive and fast timing detectors to be used in high magnetic fields and gamma detectors for positron emission tomography. The research work is supported in part by the INTAS grant 04-78-6839.

Speaker: Dr Alexey Stoykov (Paul Scherrer Institut)

14:00 Digital Signal Processing Applied to Crystal Identification in Positron Emission Tomography Dedicated to Small Animals

The recent introduction of all-digital electronic architecture in Positron Emission Tomography (PET) scanners enables new paradigms to be explored for extracting relevant information from the detector signals, such as energy, time and crystal identification. The LabTEP(tm) small animal scanner, which implements free-running 45-MHz sampling directly at the output of the charge sensitive preamplifiers, provides an excellent platform to test such advanced digital algorithms. A real- time identification method, based on an Auto-Regressive Moving-Average (ARMA) scheme, was tested for discriminating between LYSO (tr ~ 40 ns) and LGSO (tr ~ 65 ns) scintillators in phoswich detectors coupled to a single Avalanche Photodiode (APD). The algorithm, that was implemented in a 16-bit fixed point DSP from Texas Instruments running at 500 MHz, can process a sustained rate up to 640 000 events/second. Even with a low energy threshold of 250 keV applied individually, error rates < 0,4% for LYSO and < 0,6% for LGSO can be achieved with this algorithm, as compared to > 10% with conventional analog pulse shape discrimination techniques. Such digital crystal identification techniques can be readily implemented with phoswich detectors for improving spatial resolution in PET, either by increasing crystal pixellisation or by mitigating parallax errors through depth- of-interaction determination.

Speaker: Mr Rejean Fontaine (Université de Sherbrooke)

14:00 Digital signal processing for X-ray spectroscopy with semiconductor detectors at high count rate

Pile-up distortion is a major drawback in X-ray spectroscopy at high count rate. Signal processing can help to overcome this problem. Pulse width narrowing with shaping tecniques can lead to reduction of pile-up distortion; however, a low shaping time reduces the noise filtration and leads to a poor energy resolution. Hence the optimum shaping time is a compromise between noise and pile-up requirements. In this work a digital signal processing sytem for X-ray spectrometers operating at high count rate is described. The system processes the output signal of a Charge Sensitive Preamplifier (CSPA) connected to a X-ray semiconductor detector. The output signal is registered by a high speed ADC (sampling rate up to 100 MHz, 14 bit resolution) and a fully digital shaping is carried out off-line by a dedicated software instead of traditional analog electronics. The software calculates the incident photon count and the energy spectrum as well. The system can deal with count rate up to 150,000 counts/s. The benefits of the fully digital off-line shaping respect to the traditional analog shaping are emphasized. Using analog systems, shaping time tuning requires hardware manipulation. Otherwise, digital signal processing allows the shaping time to changes by simply acting on software settings, reducing costs and improving ease of operation. The system, described in this paper, carries out the complete signal processing and analisys as well (filtering, pole-zero cancellation, peak detection, pile-up detection and correction). The acquisition and signal processing software has been developed using LABVIEW platform. The software is characterized by a user friendly GUI to makes easy parameter adjustement and software setting. Tests with simulated and directly measured signals show a good system ability

Speakers: Dr Angelo La Manna (Università di Palermo Dipartimento di Fisica e Tecnologie Relative), Prof. Gaetano Gerardi (Università di Palemo Dipartimento di Fisica e Tecnologie Relative)

14:00 Efficiency of Lu2SiO5:Ce (LSO) powder phosphor as X-ray to light converter under mammographic imaging conditions.

The aim of the present study was to examine the light emission efficiency of Lu2SiO5:Ce (LSO) powder scintillator under X-ray mammographic imaging conditions. Powder LSO scintillator has never been used in X-ray imaging. For the purposes of the present study a 25 mg/cm2 thick scintillating screen was prepared in our laboratory, by sedimentation of Lu2SiO5:Ce powder. Absolute luminescence efficiency measurements were performed within the range of X-ray tube voltages (22-49 kVp) used in mammographic applications. Parameters related to X-ray detection, i.e. the energy absorption efficiency and the quantum detective efficiency were calculated. A theoretical model, describing radiation and light transfer, was employed to fit experimental data and to estimate values of the intrinsic conversion efficiency and the light attenuation coefficients of the screen. The spectral compatibility of the LSO powder scintillator to mammographic x-ray films and to various electronic optical detectors was determined by performing light emission spectrum measurements and by taking into account the spectral sensitivity of the optical detectors. Results showed that Lu2SiO5:Ce powder scintillator performed adequately well in the mammographic energy range.

Speaker: Mr Stratos David (Department of Medical Physics, University of Patras, Patra, Greece)

14:00 Growth and characterization of fast scintillator ZnO single crystal

Zinc oxide (ZnO) single crystal has attracted much attention due to its potential application in ultra violet (UV) and blue light-emitting devices. In addition, ZnO crystal is a promising fast scintillator. It is reported that doped ZnO crystals exhibited pulse rise-times <100ps and fall times <1ns, which are faster than any currently available scintillators. However, it is difficult to obtain ZnO bulk crystal due to its strong polar surfaces and volatilization at higher temperatures. In the present work, a flux Bridgman technique was developed and doped ZnO bulk crystals have been grown from a high temperature solution of ZnO-PbF2 system. The growth defects were characterized and the physical properties were measured.

Speaker: Dr weidong Xiang (College of Applied Technology, Wenzhou University)

14:00 High speed readout fot H8500 PSPMT

In this paper we present a new readout system for the H8500 flat panel photomultiplier. The system is based on the MAX1320 Analog to digital converter and reads all the 64 anodic signal simultaneosly at a rate that approaches the 250 Ksample/second typical of MAX1320 with a resolution of 14 bits. The data are read throug an FPGA by an on Chip Computer, the ETRAX 100LX, and either transmitted via Ethernet or stored on a USB mass storage device.

Speaker: Dr Valentino Orsolini Cencelli (NFN - Sezione di Roma III)

14:00 Image quality of a pixellated GaAs X-ray detector

X-ray detectors require large Z materials in order to absorb X-ray radiations sufficiently. In case of X-ray imaging, fluorescence becomes one of the limiting factors for spatial resolution and contrast at X-ray energies above the ka threshold (typical average values are 10 keV, 25 keV and 30 keV for GaAs, CdTe and CsI, respectively). Since both the energy and yield of the fluorescence for a given material increase with its atomic number Z, there is an optimum value of Z. We selected GaAs corresponding to this optimum, which, at the present time, can be epitaxially grown in self-supported thick layers to fulfil the requirements (in particular the homogeneity of the electronic properties) for imaging. Image properties obtained with this material have been evaluated and experimentally measured in terms of Line Spread Function and Modulation Transfer Function (MTF), both in cases of high and low contrast, and compared with CdTe and CsI. The spatial frequencies (lp/mm) are 7.5, 3.0 and 1.0 for GaAs, CdTe and CsI, respectively, for a MTF amplitude of 0.5, a pixel size of 50 µm and an energy of 40 keV. The image contrast obtained for a given object contrast is better for GaAs than for CdTe and CsI, at low energy and for identical thicknesses of the materials. For instance, at 32 keV, the image contrasts are 0.92, 0.65 and 0.4 for an object contrast of 1 and 0.16, 0.09 and 0.02 for an object contrast of 0.5 for GaAs, CdTe and CsI (200 µm thick), respectively. We shall describe the dependences of the image contrast versus pixel size and energy. Finally, we shall discuss the minimum detectable object size in the conditions of mammography for these materials and demonstrate that an object of a given size can be detected using GaAs with a dose at least 100 times less than using CsI.

Speaker: Dr Jacques Bourgoin (GESEC R&D)

14:00 Investigation of an In-situ Position Calibration Method for Continuous Crystal Based PET Detectors

The absence of very small crystal pixels in monolithic scintillation detectors has a number of potential advantages such as higher sensitivity, better energy resolution and continuous coordinates. In such detectors, the incidence position of the 511 keV photons on the detector surface is derived from the measured scintillation light distribution. To extract this information, we used artificial neural networks To this end, each detector module has to be position-calibrated by training the neural networks. When a neural network is trained for a specific incidence angle, it yields immediately a DOI corrected incidence position of the impinging photon. An automated procedure to simultaneously obtain the calibration data to train all the neural networks for all detector modules in a fully assembled PET system has been developed and evaluated on a simulator set-up. After calibration, images of a point sources at various radial distances were taken to evaluate the quality of the procedure.

Speaker: Mr Cedric Lemaître (Vrije Universiteit Brussel)

14:00 LHCb calorimeters HV system

The calorimeter system in LHCb aims to identify electrons, photons and hadrons. The system plays also a major role for the first level triggers. All calorimeters are equipped with Hamamatsu photo tubes as devices for light to signal conversion. Eight thousand R7899-20 tubes are used for electromagnetic and hadronic calorimeters and two hundred 64 channels multi-anode R7600 -00-M64 for Scintillator-Pad/Preshower detectors. In the same time in Molecular Imaging applications the similar photodetectors are widely used too. Requirements for the calorimeter HV system • Individual control of each detector channel; • Minimizing the number of the HV cables is desired; • The system has to operate in a radiation hard environment; • Minimizing power consumption is important; • Easy connection to Experimental Control System (ECS); • Optimisation of the overall cost. The calorimeter HV system is based on a Cockroft Walton (CW) voltage converter and distributed on detector control boards connected to ECS by the SPECS serial bus. For the ECAL and HCAL the base of each photomultiplier tube (PMT) is built with a high voltage converter and constructed on individual printed circuit board, using compact surface mount components. The base is attached directly on the PMT. There are no HV cables in the system. A Field Programmable Gate Array (rad-tolerant FPGA) is used as an interface between the ECS and the 200 control channels which are supplied by each high voltage control signal distribution board. The motherboard contains an FPGA with a SPECS interface, which is connected to the Actel HV Control FPGA via a parallel bus. The role of the HV Control FPGA is to distribute signals from the SPECS bus to the electronics mounted on the motherboard and control voltage generation daughterboards. The aim of the presentation is to describe the HV system architecture, some technical details of the electronics implementation and summarize the system performance. The presented safety and very low power consumption HV electronic system for the photomultiplier tubes can be used for various biomedical apparatus.

Speaker: Anatoli Konoplyannikov (CERN, Geneva, Switserland)

14:00 Multipixel Geiger-mode photon detectors for ultra-weak light sources.

We developed and tested prototype arrays of Single Photon Avalanche Detectors (SPAD), fabricated in silicon planar technology and working at low voltage (» 30 V) in Geiger mode operation, to achieve a photon-resolving operational mode; the size of the elementary cells is from 20 µm up to 100 µm. Coupling such a multipixel sensor to light sources as scintillators and laser, we have characterized the timing, amplitude, cross-talk, afterpulsing and the dependence from the temperature. For 50 µm diameter SPADs the measured gain is of the order of 106÷107, with dark counting rates at room temperature below 1 kHz. The detection efficiency is typically around 50% at 550 nm, 10% at 850 nm and 3% at 1000 nm. We proved that in the single photon regime, obtained with a very low laser intensity, this photodetector can reconstruct the charge spectrum and the time profile of a sub- nanosecond optical pulse, while its intrinsic time resolution is below 100 ps. Between adjacent pixels we observed a cross-talk probability of the order of 10-4, while we did not observe any indication of a dependance on the distance between the triggering pixel and its neighbour. We also tested a sensor where each pixel was optically isolated from the neighbours by means of an optical trench suitably placed around it. No difference was observed with respect to the standard sensor, and this excludes the optical cross-talk. In the meanwhile, with the Deep Lithography with Ions (DLI) technology we are working to develop micro-mechanical and micro-optical components, finalized to optimize the coupling between light sources (fibers, bulk scintillators, etc) and the multipixel sensor. The goal is to fabricate compact high-sensitivity and high- resolution solid state photon detectors, for applications in several scientific areas, in particular regarding medical diagnostics, bioluminescence, TCSPS microscopy, chemical analysis, etc.

Speaker: A. Campisi (INFN-Laboratori Nazionali del Sud, Via Santa Sofia 64, 95125 Catania Italy)

14:00 New developments in photomultiplier for nuclear medicine

In this presentation, we will develop two main aspects of the developments cureently being held within the Photonis group. We will first describe our new Time Of Flight - PET photomultipliers and discuss their results obtained with several type of scintillators (standards and new fast crystals). We will then present our new developements in multipixel detection : Planacon and XP9100-16 PMT.

Speaker: Ms Carole MARMONIER (Photonis)

14:00 Optical and scintillation properties of LuAP:Ce with intentional impurities

Optical properties (absorption in UV, slope of the optical edge, transmission and color centers induced by UV-light or intentional doping) of LuAP:Ce, LuAP:Hf and LuAP:(Ce, Hf) crystals were studied to establish compositions leading to the smallest absorption at 260 nm and the highest transparency in the range of emission. A low underlying absorption attained in Bridgman LuAP:Ce crystals enabled to follow fine variations induced by Hf, when taken even in trace (20 ppm) amounts. Optical observation of the trap ensemble modification evidenced by absorption intensity redistribution over different color centers (which is strongly dependent on Hf concentration) is discussed in terms of possible action of Hf4+ ions in the perovskite lattice. The results of scintillation studies of LuAP:(Ce, Hf), in comparison to LuAP:Ce, will include measurements of the light yield, decay time constants and other basic parameters. In addition, as part of a continuing study on spectroscopic properties of LuAP, including the red luminescence, the results of characterization of single crystals grown from extremely pure oxides (both lutetium and aluminum) will be presented.

Speaker: Dr Ashot Petrosyan (Institute for Physical Research, 378410 Ashtarak-2, Armenia)

14:00 Optical current spectroscopy of GaAs based detectors

GaAs detectors for registration of X-radiation and gamma radiation are found applications in medicine and for the tasks of high-energy physics. Since the registered signals are small enough the main problem of such devices can be formulated as follows - realization of maximal detector volume keeping constant low noise level. The capacity and return current should be minimized. The is achieved via (i) fabrication of GaAs layers with low amount of defects and low carrier concentration, and (ii) structure optimization and creation of contacts with low resistance. The non-steady-state photoelectromotive force effect was used for characterization of transport parameters of ultra-pure GaAs thin films grown on semi-insulating gallium arsenide wafers using gas-phase epitaxy technique. Such structures are used for fabrication GaAs detectors for registration of X-radiation and gamma radiation. The mechanism responsible for the effect can be described as follows. Illumination of a photoconductive sample by an interference pattern formed by two coherent light beams produces a nonuniform excitation of free carriers. Diffusion of the photoexcited carriers towards the dark regions leads to charge redistribution between traps. A space charge field grating arises; this grating is spatially shifted by 90 degrees relative to the optical interference pattern and photoconductivity distribution. Small vibrations of the light along the grating vector excites an alternating current through the short-circuited crystal because of the time-dependent phase shift between oscillating spatially-periodic free carriers and fixed space charge field distributions. The structures were fabricated at A.F. Ioffe Physico-Technical Institute (laboratory of Yu.V. Zhilyaev). For our experiments we choose the GaAs sample with the layer thickness of 400 microns. The experiments can be carried out only in the geometry of Michelson interferometer at the illumination wavelength of 532 nm. The modulation frequency was 1 kHz and the light power of the signal and reference beam on the sample’s surface was about 20 mW, the inter-electrode distance was 2,5 mm. The dependence of the signal photocurrent amplitude versus spatial frequency of the interference pattern was measured. It is important to point out that the signal peaks for the spatial frequency of the interference pattern K equal to the inverse diffusion length of photocarriers. The value of diffusion length of the as grown structure was estimated to be 40 microns.

Speaker: Prof. Igor Sokolov (Deputy head of laboratory)

14:00 Optical filter on the base of planar and fiber heterostructures of LiF and NaF compounds

Optical filters of vision, IR- and UV-regions are found of wide application for science investigation systems, for medical diagnostic equipments, for chemical analytic equipments, for thermal imaging systems, for security systems and sensors of burglar alarm systems and so on. The base materials for creation of modern optical filters are thin organic polymer films. The main disadvantages of such materials are low mechanical properties (stability of abrasion), low radiation and thermal stabilities and high cost. In present work the new inorganic optical filters on the base of planar and fiber heterostructures are proposed. The creation of heterostructures is made by means of irradiation of LiF or NaF crystals by ion beams with special procedure. As a result thin surface layers of crystals (bulk or fiber) are modified and become opaque for light. We can obtain optical filters for visible, IR or UV-regions depending on irradiation procedure. It is made due to control forming of different color centers (F, F2, F2+, F2–, F3+) in compliance with model of defect evolution under ion irradiation, which is proposed by us too. The cost of suggested filter is much lower than cost of polymer filters. Moreover, suggested inorganic filters have higher stability of abrasion and higher radiation and thermal stabilities. The methods of creation of heterostructures and optical filter are protected by two applications for RUS-patents #2005113587 and #2004123343 (patent holder is Ural State Technical University–UPI (Russia, Ekaterinburg)). We hope our invention will come useful for humanity.

Speaker: Dr Boris Shulgin (Professor)

14:00 Overview of Solid State Photodetectors for both CT and PET applications

New semiconductor detectors have recently gained a lot of attention for medical applications in general. Advances in semiconductor detector arrays made of CdZnTe or CdTe may improve both energy resolution and spatial resolution of clinical x-ray systems. Alternative system designs based on TFA technology combining photodetector arrays with CMOS electronics open a possibility for compact imaging cameras. This scenario allows for the use of alternative materials such as a-Si:H and HgI2 that can be applied alone or integrated with scintillators. We present an overview of the potential photodetector and photoconductor candidates for both computed tomography (CT) and positron emission tomography (PET) applications. Preliminary experimental results obtained with the materials mentioned above will be shown.

Speaker: Dr Danielle Moraes (CERN)

14:00 Performance of a Prototype 32-channel Preamplifier Module of the Double-Sided Silicon Strip Detector Used as the Scatterer for a Compton Camera

We are currently developing a Compton camera for applications in nuclear medical and molecular imaging. A Compton camera uses two sets of radiation detectors for which incident gamma rays are Compton-scattered in the first detector and the scattered gamma rays are then fully stopped via the photoelectric effect in the second detector. Interaction points in both detectors provide the scattered gamma- ray direction. According to Compton-scattering kinematics, a gamma-ray source can be fixed to a surface of a cone with an apex at the interaction point in the first detector, an opening Compton-scattering angle and an axis along the scattered gamma- ray track. The intersections of such cones are defined to be the three-dimensional distribution of the gamma-ray source. Our Compton camera consists of a double-sided silicon strip detector (DSSD) as a scatterer for the first detector and a 25-segmented germanium detector (25-SEGD) as an absorber for the second detector. The DSSD, operated at the room temperature, was a planar-type, high-purity, n-type silicon with active areas of 5 cm x 5 cm and thicknesses of 1.5 mm, consisting of 16 x 16 orthogonal strips on both faces. The 25-SEGD was a planar-type, high-purity, p-type germanium crystal with dimensions of 5 cm x 5 cm x 2 cm. It was segmented into twenty-five segments on the front face of a germanium crystal. Each segment, with areal dimensions of 1 cm x 1 cm, acts as an individual gamma-ray detector. The unsegmented electrode on the rear face provided energy spectra with better energy resolutions. In order to improve the energy resolution of scatterer, we have developed a prototype 32-channel preamplifier module for the low-noise performance of the DSSD. The energy uncertainty of the recoiled electron in the DSSD is attributed to the energy resolution of a detector which affects not only the resolution of the Compton scattering angle but also the threshold setting for background rejection. Performance of the DSSD was tested with 241Am (60 keV), 133Ba (356 keV) and 137Cs (662 keV) standard gamma-ray sources in a self-triggering mode and also in coincidence with the 25-SEGD. Gamma-ray images reconstructed by a simple back projection method will be presented.

Speaker: Mr Nam Young Kim (Chung-Ang University)

14:00 SiAPD Readout for LaBr3:Ce Scintillators

LaBr3:Ce represents a very promising scintillation crystal for spectrometric and imaging devices, thanks to its good scintillation properties as the high light yield (61000 photons per MeV) which allows excellent energy resolution values (3% FWHM at 662keV). It is attracting the scientific community for the potential improvement of PET and SPECT instrumentation. First measurements of energy resolution values obtained coupling large crystal samples (from 0.5 to 1 inch diameter) to PMTs show unexpected worse values than NaI:Tl for photon energies less than 100keV. On the contrary values at higher energies are close to the limit of statistical PMT contribution. In addition the scintillation light detection by standard PMT is affected by the high dynode current due to the fast and intense scintillation pulses from LaBr3:Ce (about ten times higher than NaI:Tl). The recent availability of large area Si-APDs (Avalanche Photodiode) and the progresses in LaBr3:Ce growing technique have suggested an additional study of light output performed by this photodetector. This coupling is also promising because of the good quantum efficiency (about 60%) of APD at the LaBr3:Ce emission wavelength (380nm), compared with that of bialkali photocathodes (about 25%). In this work we report, compare and discuss the results of measurements performed using three different scintillators, available in our laboratories. A 12.7mm diameter by 12.7mm thickness LaBr3:Ce crystal and a 2.0x2.0x10.0mm3 pixel of a YAP:Ce array have been used to evaluate the APD response at short wavelength. Results are compared to those obtained with CsI:Tl that is considered as the gold standard crystal for APD measurements. Two CsI:Tl crystals sized 9.0x9.0x10.6mm3 and 2.9x2.9x5.0mm3 are used for comparison with LaBr3:Ce and YAP:Ce respectively. Results obtained by APD and PMT readout are presented. Hamamatsu Si-APDs (5x5mm2 and 10x10mm2 active areas) are used as solid state light detectors. PMT measurements employ an Hamamatsu R6231 and a Thorn EMI 9765B05 with 2 and 3 inch photocathode diameter respectively. A study of the energy resolution has been focused in the energy range from some tens of keV to 662keV to investigate the LaBr3:Ce response at photon energies below 100keV showed by Dorenbos as the limit of better energy resolution of this scintillator with respect to the competitors. This energy range is particularly important for nuclear medicine imaging. Spectra obtained at 22°C with YAP:Ce and LaBr3:Ce crystals coupled to the 25mm2 APD show the 81keV peak from 133Ba well resolved from the low-energy noise. The energy resolution values measured at 122keV coupling the PMT to the YAP:Ce and LaBr3:Ce crystals (27.0% and 7.5% respectively) improve by about 15% when using APDs. The APD readout of CsI:Tl crystals (9.0x9.0x10.6mm3 and 2.9x2.9x5.0mm3) gives energy resolution values of 15.6% and 10.5% respectively.

Speaker: Dr Raffaele Scafè (Casaccia Research Center, ENEA, Rome, Italy and INFN, University of Rome “La Sapienza”, Rome, Italy)

14:00 SiAPD Specifications for Multi-face Scintillation Readout

This work is aimed to design a gamma-ray scintillation detector, suitable for radionuclide imaging and able to measure event-by-event the energy and the point-of- interaction of primary radiation. In principle, to obtain these quantities, multi- face readout of a parallelepiped crystal is necessary. Light intensity values from each crystal face are used for centroiding the event through an appropriate algorithm, while the sum of responses is proportional to the energy of the detected gamma-ray. The efficiency of light collection is affected by some geometrical factors concerning the crystal-APD assembly. In particular the width of APD peripheral dead zones and the thickness of the APD entrance window affect the portion of scintillation light detected by the photosensors. To evaluate the influence of these factors on the assembly response, in terms of total light collection and position evaluation, a simplified model has been developed. The model calculates luminance values at photosensor’s active areas as a function of the point of interaction. It only considers geometric effects, under the hypothesis of isotropic scintillation emission, and negligible light reflection at crystal walls. This first step hypothesizes the same interaction probability in every point of the crystal while a further study takes into account, using a finite-element model, the probability of photoelectric interaction as a function of photon energy and crystal material. A detector setup has been considered including a crystal sized 11.4x11.4x12.4mm3 and 6 photosensors, each coupled to a face of the scintillator. A number of cases with the APD overall dimensions fitting the crystal face size, but having different active areas, have been considered. Further cases have been considered, with increasing optical window thicknesses. In particular, simulations concerned photosensors with rectangular active areas surrounded by dead zones in the range from 0mm to 1mm and optical windows in the range from 0 to 0.5mm. The highest considered values for both parameters are considered as easily obtainable using the technology available for detector assembling. Energy and position responses vs. dead zones width and optical window thickness are presented. The first study shows the strong influence of dead areas on both analyzed quantities, mainly at crystal borders. A dead zone width of 0.5mm and an optical window 0.3mm thick seem realistic values to obtain a response with acceptable linearity inside an inner crystal volume approximately correspondent to active zones. The second study points out how the light collection is affected by photon energy and crystal material, which determine the depth of interaction. According to this schematization the detector response can be represented, voxel-by- voxel, as the product of two parameters obtained from the studies above described. This study helps to design the detector assembly, in terms of crystal material, size and light readout geometry, to optimize its response in the fixed energy range.

Speaker: Dr Raffaele Scafè (Casaccia Research Center, ENEA, Rome, Italy)

14:00 Silicon Photomultiplier: Novel Photodetector for Medical Imaging Systems

Development of photodetectors for detecting of low intensity photon flux is one of the critical issues for experimental physics, medical imaging system and many others. Review of silicon photomultipliers SiPM, novel type of avalanche photodetector with Geiger mode operation is presented. The structure of photodetector is based on the matrices of microcells with density of 1500/mm2, working in Geiger mode with internal gain of amplification 105 – 106 and integrated quenching mechanism. Photon detection efficiency of order 20-30 % for green spectrum of light and estimated time resolution is about 30 ps. The structure of the photodetector gives the possibility of detecting the low flux up to 1500 photoelectrons with proportional responce. The physics of silicon photomultipliers, recent development and experimental test with different kind of scintillation materials will be presented. Novel type of photodetector - Silicon Photomultiplier is compact, robast, not sensitive to magnetic fields and will find wide application in medical imaging system especially in Positron Emission Tomography (PET) and combine PET imaging system.

Speaker: Prof. Valeri Saveliev (Obninsk State University)

14:00 Small Animal PET slow control system

We present our technical studies of an automated slow control system able to control a small animal PET (Positron Emission Tomography) Gantry. It consists of a cable-less powered embedded acquisition cards, that include the detectors, that are wireless and interfaced with the main data acquisition system; both connected to an event builder computer station. The gantry angle position is red out trough an opto-scanner allowing angular precision of 0.0046° for an absolute angle position coding. The Gantry, free and continuous, rotation is controlled with an embedded microcontroller. The later ensures real time interface with the global data acquisition system (daq). Individual absolute detector position can be acquired independently with an under study dedicated laser calibration system.

Speaker: Mr Julien lollierou (IPNL (Institut de Physique Nucléaire de Lyon))

14:00 Study of multipixel Geiger-mode avalanche photodiodes as a readout for PET.

Multipixel Geiger mode APDs (also known as MRS APDs, AMPDs and SiPMs) with different structures and sensitive areas (from 1x1 mm^2 to 3x3 mm^2) have been studied as a readout for LSO, LYSO and LYAP scintillator crystals. Energy and timing spectra were measured using Cs-137 and Na-22 gamma sources. The results of this study allows us to conclude that this photodetector is a very promising candidate for PET applications.

Speaker: Yuri Musienko (Northeastern University)

14:00 The electronics readout and the DAQ system of the DRAGO Anger Camera

The scope of the DRAGO project, supported by Italian INFN, is the realization of a high resolution, compact gamma ray imager, based on the Anger camera principle. In this configuration, the light generated by a unique scintillator is read by an array of 77 Silicon Drift Detectors. In order to locate the position of interaction of the photon inside the scintillator it is necessary to make an amplification and filtering of the detector signals followed by a processing of the acquired data. The electronics readout and processing system can be divided in two separate parts: the analog front end and the DAQ board. The analog front end is composed of 80 readout channels divided in 10 CMOS chips, realized in the 0.35um AMS technology, each one processing 8 channels. Each analog channel of the circuit includes a low-noise preamplifier, a 6th order semigaussian shaping amplifier with four selectable peaking times from 1.8us up to 6us, a peak stretcher and a baseline holder. The integrated time constant used for the shaping are implemented by means of a recently proposed ‘RC’ cell. This cell is based on the de-magnification of the current flowing in a resistor R by means of the use of current mirrors. The 8 analog channels of the chip are multiplexed to a single analog output. A suitable digital section provides self-resetting of the channels, trigger output and the programming of independent threshold on the analog channels by means of a programmable serial register and 3bit DACs. The energy resolution measured using a single channel of the chip with a Silicon Drift Detector Droplet (SDD3) is of 128eV at 6keV with the detector cooled at -20°C. The multiplexed outputs signals are the sent from the CMOS circuit to the acquisition system. For each gamma event, the acquisition system performs the A/D conversion of all the signals of the array and sends them to a host PC, where the position reconstruction is executed on-line. The DAQ board contains 10 ADCs, each one dedicated to a single ASIC of the analog section. When a gamma event takes place, each ADC converts the sequence of signals coming from the analog multiplexer, with a resolution of 13 bit (ENOB). The burst conversion rate of the board is 50 Ms/s and is limited essentially by the analog section; this gives a dead time of about 2us per event. The converted data are stored in a FIFO memory, for buffering, and then are transferred to the host PC via a USB 2.0 bus. The data processing for the reconstruction of the position of interaction of the event in the Anger Camera is based on a correlation filter implemented in C++.

Speaker: Alberto Gola (Politecnico di Milano and INFN Italy)

14:00 The novel high-gain PET detector for simultaneous PET/MR imaging

The design and proof-of-principle results for an MRI-compatible detector module for ‎PET are reported. The proposed detector has a ‎finely pixelated Cerium-doped Lutetium ‎Yttrium Orthosilicate (LYSO) array in optical contact with an ‎array of a- Se avalanche ‎photosensors. The quantum efficiency of the a-Se photosensor was measured as a ‎function of the wavelength of optical ‎excitation, and is ~95% at the scintillation peak of ‎LYSO. The gain of the photosensor and the ionization coefficients of electrons and holes ‎in a-Se were measured as functions of the electric field induced by a high voltage bias. ‎The photosensor can provide a gain of 10^3 and has extremely low dark current (< 1 nA). ‎When ‎combined with LYSO, a crystal with high light yield (~30 photons/keV) and fast ‎scintillation ‎decay (40 ns), it forms a PET module with high detection efficiency. The ‎insensitivity of the module’s components to ‎magnetic fields makes it a promising ‎‎candidate for the development of a PET detector for simultaneous PET/MR imaging. ‎Such a system would enable visualization of relationships between anatomical structure ‎and function, thereby greatly aiding the treatment planning and monitoring of many types ‎of cancer.‎

Speaker: Dr Alla Reznik (Sunnybrook&Women’s College Health Sciences Centre)

14:00 X-ray spectroscopy and dosimetry with a portable CdTe device

Abstract X-ray spectra and dosimetry information are very important for quality assurance in X-ray diagnostic systems. Several semiconductor detectors (Si, Ge) have been proposed for medical X-ray measurements [1,2]. Recently, cadmium telluride (CdTe) semiconductor is promising X-ray detector and is suitable for portable systems. The number of papers investigating the physical and the spectrometry properties of CdTe detectors [3] has increased many times in the course of the 90s and keeps on growing. The high atomic number (Cd: 48, Te: 52) and the high density (~ 5.9 g cm-3) of CdTe crystals ensure high detection efficiency for photons with energy below 100 keV even for thin detectors; the wide band gap (EG ~ 1.5 eV) enables CdTe detectors to be operated at room temperature without cryogenics. In this work a CdTe portable apparatus for X-ray spectroscopy and dosimetry is described. The system is able to directly measure (count mode of operation) X-ray spectra and X-ray fluence at high photon count rate, as typical of medical X-ray diagnostic systems. The portable device consists of four blocks: a detector case (detector and preamplifier), a shaping amplifier, an ADC card and a notebook computer. Schottky contacts and a thermoelectric cooling (-30° C) of both the CdTe crystal and the preamplifier input FET ensure low noise and good stability in the X-ray measurements. The output preamplifier pulses are processed by the shaping amplifier and then are recorded by a 12 bit ADC card with a 10 MHz sampling rate. A dedicated software calculates the incident photon count and the energy spectrum by analyzing the sampled output of the ADC card. Good system response to monoenergetic photons was measured using X-ray and gamma-ray calibration sources (109Cd e 241Am). Measured molybdenum X-ray tube spectra show the good spectral system ability in mammographic energy range (1-30 keV) also at high photon fluence rates (~ 106 counts/mm2 s). Dosimetry measurements have been made using radioactive sources and a molybdenum X-ray tube. The system compactness and its easy operation, in addition to good energy and time resolution, make the system suitable for medical X-ray spectroscopy under clinical conditions. Keywords: X-ray spectroscopy, dosimetry, semiconductor detectors, CdTe References [1] Birch, R., Marshall, M., Computation of bremmstrahlung x-ray spectra and comparison with spectra measured with a Ge(Li) detector, Phys. Med. Biol., 24, p. 505-517 (1979). [2] Aoki, K., Koyama, M., Measurement of diagnostic x-ray spectra using a silicon photodiode, Med. Phys., 16 (4), p. 529-536 (1989). [3] 11th International Workshop on Room Temperature Semiconductors and Associated Electronics, Nucl. Instr. and Meth. A 458 (2001) 1-603.

Speaker: Mr Angelo La Manna (DIFTER University of Palermo)

14:00 XPAD3: A new photon counting chip for X-ray CT-Scanner

The XPAD3 (X-ray Pixel chip with Adaptable Dynamics) circuit is the next generation of 2D X-ray photon counting imaging chip to be connected to a pixel sensor using the bump and flip-chip technologies. This circuit, designed in submicronic (0.25µm) IBM technology, contains 9600 pixels (130um x 130um) distributed into 80 columns of 120 elements each. Its features have been improved to provide high counting rate over 1Mphotons/pixel/s, high dynamic range over 60KeV, very low noise detection below 100e, energy window selection and fast image readout below 2ms/frame. The analog part in each pixel gathers a low noise charge sensitive preamplifier, a voltage to current converter, and current comparators. The digital part of XPAD3 is meant to count hits in each pixel and to configure, calibrate, test and readout the chip. An innovative architecture has been designed in order to prevent the digital circuits from bothering the very sensitive analog parts placed in their neighborhood. This allows to read the chip during acquisition while conserving the precise setting of the thresholds over the pixel arrays. Finally, the aim of this development is to combine several XPAD3 to form the PIXSCAN project (see poster " PIXSCAN: Pixel Detector CT-Scanner for Small Animal Imaging " at this conference). A large surface (8 x 12 cm2) X-ray detector will also be developed using the XPAD3 for crystallography (including proteins) on synchrotron facilities (ESRF, SOLEIL, France).

Speaker: Mr Patrick Pangaud (Centre de Physique des Particules de Marseille (CPPM), France)

15:00 → 18:00 Other Modalities

15:00 Ultrasound Mammography

Breast cancer is the first cause of mortality of women aged 55 to 75. It is only in the last decade that breast anatomy has started being seriously considered by pioneer radiologists such as M. Stravos, M. Teboul and D. Amy. Indeed, it is now widely admitted that 90% cancer lesions and mainly all the specific diseases of the breast initiate from the epithelium and develop first in the ductolobular structures (infiltration of Cooper's ligaments). Thus, the restitution of breast anatomy appears of prime importance since the examination of the gland should be guided by a relevant exploration strategy. We introduce a near-field formulation of the acoustic field scattered by a soft tissue organ such as the breast (assumed to be weakly heterogeneous). This derivation is based on the Huygens-Fresnel principle that describes the scattered field as the result of the interferential scheme of all the secondary spherical waves. This derivation leads us to define a new Fourier transform which yields a spectrum whose harmonic components have an elliptical spatial support. Based on these projections, we define the Elliptical Radon transform and show that it is possible to reconstruct either the impedance or the celerity maps of an acoustical model characterized in terms of impedance and celerity fluctuations. We observe that this formulation is very similar to that developed in the far field domain where the Radon transform pair is derived from an harmonic plane wave decomposition. This formulation allows us to introduce the Ductal Tomography, following the example of the Ductal Echography, that provides a systematic inspection of each mammary lobe, in order to reveal lesions at an early stage. In order to review the performances obtained with current echographs in view of specific experiment (numerical simulations), we develop a computer phantom that gains in realism. This 2-D anatomical phantom is an axial cut of the ductolobular structure corresponding to a daisy-like internal arrangement with petals (lobes) radiating around the nipple, for healthy and pathological situations. The different constitutive tissues and ducts are characterized in terms of density and celerity parameters whose spatial distributions are defined with specific random density laws. The use of a velocity-pressure formulation permits us to model time domain acoustic wave propagation. Broadband US pulses are transmitted and measured in diffraction around the breast with a ring antenna, the images are reconstructed using the elliptical back-projection-based procedure mentioned above. The results show that a tomographic approach provides much more information both about the breast structure (navigation) and about the lesion (specificity) than conventional echography. We conclude by introducing the mammograph prototype in development in our laboratory and some perspectives in molecular ultrasounds based on contrast agent microbubbles. Keywords : ultrasound mammography, near-field, anatomic breast phantom, prototype.

Speaker: Dr Serge Mensah (CNRS - LMA)

15:15 Whole body small animal examination with a diffuse optical tomography instrument.

Three dimensional Optical Diffusion Tomography allows in vivo studies of tumour life without any stress or damage for the animal. We present hereafter theoretical and experimental results obtained with the system developed in our laboratory on phantoms and mice. The experimental set-up consists of a laser source (690nm) coupled to a motorized stage, a CCD camera and a tank to receive the animal. The excitation and emission wavelengths of the system and the fluorophore are chosen to optimize transmission through the whole animal (leaver, lungs …). The acquisition geometry is considered as infinite in (x, y) directions and of thickness z varying from 10 to 15mm. For mice acquisitions, this is achieved by immersing them in an index matching medium. Reconstruction is performed through an ART algorithm based upon a fine description of material-light interaction taking into account diffusion as well as absorption phenomena. A first study on phantoms was conducted to evaluate depth resolution: two glass capillary tubes filled with Cy5 (diameter=1 mm, Length=20 mm) separated in z by various distances (axis to axis) are immersed in a diffusing medium. A depth resolution of 4mm is achieved on simulation and of 4,5mm on experimentation. A second study is performed on mice. 4 healthy mice and 13 lung metastasis bearing mice (mammary murine tumour) are imaged at different stages of the tumour development: 12, 13, or 14 days after the primary implantation. Acquisitions are made 3 hours after intravenous injection of 150 microgramme Transferine/Alexa 750. The reconstruction of the fluorophores concentration for a healthy mouse shows roughly nothing in the lungs. They are slightly visible on the volume slices showing very little fluorescence. Only a few markers concentrate in the lungs. The same reconstruction done for mice bearing tumours shows, on the contrary an accumulation of fluorophores in the lungs. A control study has been conducted on two mice bearing a 14 days old tumour that are non-injected with Transferine/Alexa 750. In this case, the reconstruction area does not present any fluorescence. These first results show our system performances for reconstructing whole body mice in slab geometry even in the area of the lungs. Detection and localization of the fluorophore fixations are presented according to the stage of the tumour development. Our results show that the system is able to separate healthy from cancerous mice. We have compared these results to the FRI signal observed on lungs of dissected mice.

Speaker: Dr Anne Koenig (LETI - CEA Recherche Technologique)

15:30 Fluorescence Tomography Enhanced by Taking into Account the Medium Heterogeneity

Fluorescence tomography modality is envisioned to be an economical functional quantitative measure and could benefit to the pharmacological industry for experiments on small animal. It consists in injecting cancer specific fluorescence marker and reconstructing the fluorescence by using the fluorescence tomography technique. The reconstructed fluorescence can be used as a marker of the cancer activity to evaluate the efficiency of treatments. Currently, the technique is limited by the large heterogeneities of biological tissues which create artefacts on the reconstruction and reduce the quality of the subsequent analysis. Here, we describe how we take these heterogeneities into account and show the efficiency of our method through a phantom experimental validation. Fluorescence tomography experiments consist in illuminating a highly scattering medium with near infrared light (excitation wavelength) and measuring the transmitted light at the excitation and the fluorescence wavelengths thanks to the use of filters placed in front of the camera. Our light source is a continuous laser (690nm, 26 mW, Powertechnology) beam coupled into a multimode optical fibre and focused by a lens to a single spot underside the studied medium. The fibre and the lens are driven by two motorized translation stages that allow XY scanning of the specimen. At each step of the laser spot, light is collected from the specimen and imaged onto the cooled CCD camera using a 25 mm f/l objective. For this experiment, we position a cylindrical tank (interior dimensions: diameter=6 mm, height=6.5mm) filled with 1micromol/L of Alexa750 fluorophore positioned at 1.5mm from the bottom of a cylindrical-shaped box (diameter=110 mm) filled up with a scattering liquid (mixture of water, black India ink as the absorber and intralipid (Fresenius Kabi) as the scattering medium). Measures at the excitation wavelength allow the determination of a map of the optical properties of the medium. As opposed to the methods classically employed, we show in this paper, that the use of Green functions G adapted to the inhomogeneous medium allows better reconstruction of the fluorescence map. The calculation of the transfer functions G for an inhomogeneous medium follows the mathematical derivations classically used in the literature. G are deduced from the difference between the measurements and the measure we would have for a homogeneous medium with known optical parameters (homogeneous absorption and reduced scattering coefficients). It can be obtained by repeating the experiment described before without the fluorescent cylindrical reservoir or by using, as we did, an analytical forward model. The grid for the attenuation and fluorescence reconstruction is a 14x14x15 mesh. We first reconstruct the attenuation map of the medium along with the G functions corrected from the medium attenuation variations. It shows the presence of a cylindrical region of low attenuation at 2.5mm<z<9.5mm, explained by the fact that no ink was added in the fluorescent solution. We then reconstruct the fluorescence map (20 iterations of ART (algebraic reconstruction technique), relaxation=0.01). It shows that the reconstruction with corrected G functions is closer to the expected values than non corrected reconstruction especially when the z resolution is considered. Vertical profiles of the two distributions in the axis of the fluorophore tank shows the z-localization and z-resolution are better when the reconstruction is performed with corrected transfer functions. To obtain more accurate fluorescence tomography reconstruction, we recommend the use of G functions corrected from variations of the optical properties of the studied object.

Speaker: Dr Lionel Hervé (LETI - CEA Recherche Technologique)

15:45 Feasibility of performing whole body micro-CT mouse imaging in less than a minute

Recent developments in the field of micro-CT imaging have revolutionized the ability to examine in vivo the living experimental animal models such as mouse with acceptable spatial resolution according to the animal size. One of the main requirements of in vivo imaging for biological researchers is a reduced acquisition and reconstruction time for screening purposes. We previously introduced inline acquisition and reconstruction architecture to obtain in real time the 3D attenuation map of the animal [1]. The micro-CT system, component of the AMISSA platform, is based on commercially available X-ray detector and micro-focus X-ray source. The reconstruction architecture is based on a cluster of PCs where a dedicated data communication scheme combining serial and parallel treatments is implemented. One of the key points is to take full advantage of hyperthreading capabilities proposed by recent processors. A dedicated data acquisition system is also developed to obtain high performance transmission rate between the detector and the reconstruction architecture. With the proposed architecture, we demonstrated that is already possible to obtain a real-time reconstructed image of a whole body mouse in 358s. This value corresponds to the time required to acquire 768 projections of 2048x2048 pixels and to backproject them inside a volume of 140 Mega voxels with 100µm spatial resolution. In order to decrease this acquisition/reconstruction time, several parameters have to be optimized such as the binning of the projection and the number of projections. The aim of this abstract is to demonstrate the possibility to reach an acquisition/reconstruction procedure in less than 20s for the entire body of the mouse fulfilling the requirements of in-vivo imaging. [1] Brasse D et al, Towards an Inline Reconstruction Architecture for micro-CT Systems, Phys. Med. Biol. 50 (2005) 5799-5811.

Speaker: Mr David Brasse (Institut Pluridisciplinaire Hubert Curien, UMR 7178, Strasbourg, France)

16:00 The X-Ray CT Component of the NanoSPECT/CT Small-Animal Imaging System

In previous works we have presented our multiplexing multi-pinhole SPECT imaging technique and its extensive applications in small-animal molecular imaging. SPECT combined with X-ray CT introduces anatomical information and improves acquisition (helps define axial region of interest), reconstruction (attenuation correction) and data analysis (aids segmentation). In this work we present a description of an X-ray CT upgrade to a dedicated small-animal SPECT system (the NanoSPECT). The NanoSPECT houses up to four gamma cameras outfitted with multi-pinhole apertures providing submillimeter SPECT resolution. The X-ray source and detector are mounted on the back of the high-precision gantry and thus share the same axis of rotation as the SPECT system. Helical scanning is employed by both modalities and is performed by translating of the animal through the SPECT and CT fields of views. The system is capable of acquiring partial- or full-body mouse and rat images ranging from 40 to 270mm. This variable axial-length feature is also present in the SPECT modality. The X-ray source is a 90kVp microfocus (18µm) tube. The X-ray detector is made up of a 1024x2048 array of 48µm pixels (49.2x98.6mm2) and reads out at a rate of 2.7fps. The geometric magnification of the system is 1.3 providing a reconstructed CT resolution below 150µm (below 1.0 mm for SPECT). Reconstructions are performed using a ray-tracing based filtered backprojection and the system is setup for image acquisitions ranging from quick low-dose to high-resolution studies. We will present a wide range of dual-modality phantom and animal studies.

Speaker: Dr Christian Lackas (Research Center Juelich)

16:15 break

16:30 Coffea break

17:00 3D-imaging using micro-PIXE

In Particle Induced X-ray Emission (PIXE), the production cross sections of characteristic X-rays are very large in comparison to those of continuous X-rays, which when integrated over the detector resolution are typically 3 orders of magnitude smaller. This feature is quite different from the case of electron bombardment, where the electron bremsstrahlung contributes predominantly to the X- ray spectrum. Due to this inherent advantage, X-rays from a pure metal target bombarded with micro-beams can be considered to constitute a monochromatic X-ray source. We use this feature in an X-ray CT, which then provides the 3 dimensional structure of a small object with a resolving power of micron size. On the basis of this idea, we develop 3D imaging consisting of a micro-beam system and an X-ray CCD camera. The performance of the X-ray CCD camera( Hamamatsu photonics C8800X) is as follows : element size is 8mm 8mm, and the total number of image elements 1000 1000 provides an image size 8mm 8mm. A biological sample is placed in a tube with inside diameter 1000mm and wall thickness 25mm. This tube is rotated by a stepping motor and 2D transmission images of the sample are taken with characteristic K-X-rays produced from a metal target bombarded by 3MeV proton micro-beams. 3D images are reconstructed from these 2D projection images by using an iteration method. We applied our system to investigate an in-vivo sample, in this case being a very small ant. The ant was anaesthetized with chloroform and exposed to characteristic X-rays of Ti. The 3D image of the ant was obtained with a spatial resolution of 4mm. In the usual X-ray CT, X-rays of ~60keV are used and the images reflecting the density of object are obtained. In our case, due to absorption edge and the strong dependence on atomic number of photo ionization cross sections, we obtain images emphasizing the contribution from heavier element. The distribution of K in the gnathic glandula could be clearly imaged. It is expected that our 3D imaging system could provide cross sectional images of in-vivo samples with high spatial resolution and may thus be applied to a wide range of researches in biology and medicine.

Speaker: Prof. KEIZO ISHII (TOHOKU UNIVERSITY, SCHOOL OF ENGINEERING)

17:15 Phase contrast enhanced high resolution X–ray imaging and tomography of soft tissue

Imaging of soft tissue demands high sensitivity and contrast enhancing techniques in order to reliably detect and resolve structures and patterns of small size and of similar atomic composition. For these purposes, traditional X–ray transmission–based radiography is limited by the low and homogenous absorption of X–rays in soft tissue. Phase shift contrast imaging can bring additional information by diffraction and interference effects in the wave–nature properties of photon beams induced by varied media and structures in soft tissue. We investigate the potential and advantages of phase contrast as a supplementary tool in X-ray imaging and tomography of soft tissue in small insects and organic samples. We implement high spatial resolution soft X–ray 2D and 3D imaging making use of the state–of–the–art Medipix–2 semiconductor pixel detector and microspot X–ray tube in addition to geometric magnification. We achieve high performance soft X–ray detection by our single photon counting system having high sensitivity and unlimited dynamic range. Data collection has been eased by a new USB–based readout together with online image generation. Improved signal reconstruction techniques make use of advanced statistical data analysis, enhanced beam hardening correction using individual pixel calibration and devoted algorithms for tomographic reconstruction.

Speaker: Mr Carlos Granja (researcher)

17:30 MAGNETIC RESONANCE IMAGE RECONSTRUCTION USING ANALYTIC IMAGE REPRESENTATION

Partial k-space acquisition is a subject of great interest in MRI. This type of acquisition usually consists of, in addition to the positive half of k -space, acquiring a number of negative frequencies beyond the k-space center, thus requiring an increase of acquisition time. Furthermore, the quality of the reconstructed image depends on the number of negative frequencies acquired. To date it has not been demonstrated that the non acquisition of negative frequencies of k-space can produce useful reconstructed images. We propose a new approach for reconstructing MR images from partial k-space. It is based on the notion of analytic image and presents the particularity of using only exactly half of k-space without involving any negative frequencies. We first give a systematic formulation of the notion of analytic image, by defining that an analytic image associated to a real image is a complex image whose real part is equal to the real image to be reconstructed, and whose imaginary part is the two- dimensional Hilbert transform of the real part. Equivalently, in k-space, the Fourier transform (FT) of an analytic image is zero for the negative (or positive) half of k-space, and twice the FT of the analytic image’s real part for the positive frequencies. So, in practice, given a half of k-space, we first consider the omitted half k-space as zero, and multiply the acquired k-space samples by a factor of 2. Then, the thus obtained k-space is inverse Fourier transformed, and the image is reconstructed by taking the magnitude of the complex image. The proposed MR reconstruction method was evaluated on both simulated and real MR data. For simulated data, a slowly varying phase was introduced to generate a complex image whose k-space does not exhibit Hermitian symmetry. For the real MR data, the raw k-space data were acquired using a GE 1.5 T MRI system. The reconstruction quality was assessed both visually and using quantitative criteria such as root mean square (RMS). The obtained results showed that, with the conventional conjugate symmetry method (without using negative frequencies as well as phase estimation), the reconstruction quality depends on the type of images and characteristics of the undesired phase shifting, and is characterized by the presence of important artefacts as reported in the literature, whereas with the proposed method, the reconstruction quality is fairly good and stable. Moreover, in all cases, the proposed method exhibited a substantially better reconstruction quality than the conjugate symmetry method. So, the proposed method based on analytic image representation offers the possibility of reconstructing acceptable images using only one and exactly one half of k-space without involving any negative frequencies, or phase correction.

Speaker: Ms Josiane YANKAM NJIWA (CREATIS)

17:45 TomXgam: development of a combined TEP/CT scanner for small animals

Recently, the development of combined TEP/CT imaging systems led to a rapid expansion of this technique in clinical routine. Like for clinical routine, the development of dedicated PET scanners and micro- CT scanners for small animals encourages to have these two imaging modalities embedded on a common gantry. However, the juxtaposition of both modalities does not allow for extrapolating exact position of the animal during the PET scan. Therefore, we intend to combine phoswich detector modules developed for the ClearPET small animal PET scanners with hybrid pixel detectors developed for the small animal PIXSCAN micro-CT scanner in a common rotating gantry to study simultaneous TEP/CT hybrid imaging. Ultimately, the use of hybrid X and gamma detector modules capable to detect simultaneously X rays and gamma rays would allow for developing new methods to correct PET data for the animal movements due to respiration or cardiac motion by using transmission data correlated with physiologic signals.

Speaker: Christian Morel (Centre de Physique des Particules de Marseille (CPPM))

18:00 → 19:00 Round Table Cerimed

Thursday 11 May

08:30 → 12:40 Clinical Imaging

08:30 Advanced molecular imaging techniques in the detection, diagnosis, therapy, and follow-up of prostate cancer

Prostate cancer is the most frequently occurring cancer in men. Considerable improvements have been implemented in the diagnosis with the Magnetic Resonance Imaging (MRI) technique, and with nuclear medicine techniques of scintigraphy and Positron Emission Tomography (PET). Unfortunately, the presently used standard clinical nuclear medicine imagers for single gamma imaging (gamma cameras) and for positron prostate imaging (PET scanners) are not optimized for the task. The sensitivity, spatial resolution and lesion contrast attained are inferior to what can be potentially achievable with optimized dedicated prostate imagers and procedures. Following the successful example of a Conference organized in Rome in 2001 on breast imaging, a Symposium took place in Rome (December 6-7) to investigate what kind of complementary value can be provided by dedicated nuclear medicine molecular imaging techniques to the MRI or ultrasound imaging of prostate which is primarily focusing on the structural information, while the functional metabolic or molecular imaging attained with nuclear medicine modalities can offer adjunct information about the stage of the disease and can be of potential important assistance during the therapy (chemo and radio) and follow-up phases of the disease. New technologies, mainly involving gamma ray detection and high resolution scintigraphy are being studied to overcome limitations of the standard single gamma imaging. For example, Compton imaging of prostate offering in principle high efficiency of signal detection from the uptake in this organ, is easier to introduce than Compton imaging of larger organs. New radiopharmaceuticals are being studied to improve detection sensitivity of single gamma and PET modalities and to detect in-vivo biological characteristics of prostate cancer, in order to guide medical treatment of the disease. Only combined advances in instrumentation, radiopharmaceuticals, and in imaging and treatment procedures will lead to much better diagnosis and treatment of prostate cancer and ultimately will save many lives. During the Symposium the knowledge status of prostate cancer biology, as well as the methods of prostate cancer diagnosis were reviewed. The limitations of presently utilized imaging techniques (structural and functional) in prostate cancer screening, detection, and treatment were critically reviewed. The major therapeutical techniques (surgical, medical as well as radioterapeutical) were discussed. A satellite Technical Workshop on New Nuclear Medicine Detectors For Imaging Prostate Cancer took place at the end of the Symposium to review the most recent progress in the radionuclide imaging detector technologies.

Speaker: Garibaldi Franco (Italian Institute of Health and INFN - Roma1, gr. Sanita')

09:00 Molecular Breast Imaging: first results from Italian National Health Institute clinical trials

Dedicated high resolution detectors are needed for detection of small tumors by molecular imaging with radionuclides. Absorptive collimation typically used for imaging single photon emitters, results in a strong reduction in efficiency. System based on electronic collimation is offering higher efficiency but is complex and expensive. For these reasons, the simulations and measurements have been performed to design dedicated high resolution mini gamma cameras. Critical parameters are contrast and Signal to Noise Ratio (SNR). Intrinsic performance (spatial resolution and pixel identification, linearity) were optimized as first. Pixellated scintillator arrays (NaI(Tl)) of different pixel sizes were coupled to arrays of PSPMT’s with different anode dimensions (6 x 6 mm2 and 3 x 3 mm2). Detectors having a Field Of View (FOV) of 100 x 100 mm2 and 150 x 200 mm2 were designed and built. The electronic system allows readout of all anode pad signals. The collimation technique was then considered and the limits of the coded aperture option were studied. First clinical trials performed in the framework of the Italian National Insitute project and Universities of Rome Tor Vergata and Napoli will be presented.

Speaker: Dr Francesco Cusanno (Istituto Nazionale Fisica Nucleare)

09:15 ClearPEM: a PET imaging system dedicated to breast cancer diagnostics

The Clear-PEM scanner for positron emission mammography under development is described. The detector is based on pixelized LYSO crystals optically coupled to avalanche photodiodes (APD) and readout by a fast low-noise electronic system. A dedicated digital trigger and data acquisition system is used for on-line selection of coincidence events with high efficiency, large bandwidth and negligible dead-time. A specialized gantry allows to perform exams of the breast and of the axilla. In this paper we present the imaging performance estimated from Monte Carlo simulated data and the first results of the quality control of the detector modules that integrate the system under construction.

Speaker: Prof. Joao Varela (LIP)

09:30 Image reconstruction techniques evaluation and validation for high resolution human brain PET imaging

High resolution PET imaging is now a well established technique not only for small animal, but also for human brain studies. The ECAT HRRT brain PET scanner (Siemens Molecular Imaging) is characterized by an effective isotropic spatial resolution of 2.5 mm, about a factor of 2 better than for state-of-the-art whole-body clinical PET scanners. Although the absolute sensitivity of the HRRT (6.5 %) for a point source in the center of the field-of-view is increased relative to whole-body scanners (typically 4.5 %) thanks to a larger co-polar aperture, the sensitivity in terms of volumetric resolution (75 µm3 at best for whole-body scanners and 16 µm3 for the HRRT) is much lower. This constraint has an impact on the performance of image reconstruction techniques, in particular for dynamic studies. Standard reconstruction methods used with clinical whole-body PET scanners are not optimal for this application. Specific methods had to be developed, based on fully 3D iterative techniques. In this study, we present the performance of several implementations of the fully 3D sinogram-based OSEM (Ordered Subset Expectation Maximization) image reconstruction algorithm, with different noise (weighting scheme) and spatial resolution models. Their performance was evaluated for the neuronal dopamine transporter (DAT) imaging in humans with the HRRT, using a selective DAT radioligand ([11C]-PE2I). Eleven healthy volunteers were scanned, and the striatum binding potential (BP) of [11C]-PE2I was calculated using a simplified reference model with the cerebellum as the non-specific binding region. Another set of eleven healthy volunteers was scanned on the whole-body ECAT EXACT HR+ scanner (Siemens Molecular Imaging) using the same protocol. The BP values obtained on the HR+ with a well validated analytic reconstruction technique (3DRP) were considered as the reference. The BP values obtained with the HRRT were validated against the HR+ reference values by post-processing the HRRT images to match the partial volume effect of the HR+ images. The use of a simplified noise model in the HRRT reconstruction algorithm introduces a systematic quantitative bias in the activity measured in the image when the relative amount of random coincidences is high, for example at the beginning of a dynamic acquisition. The modeling of a realistic intrinsic spatial resolution allows noise amplification in the image to be reduced, in particular at very low count levels (for frame duration of the order of one minute or later frames for 11C labeled radioligands). These results validate the use of the HRRT for cerebral DAT imaging in humans, and show the importance of using a realistic modeling of the acquisition process in the reconstruction algorithm for high resolution dynamic brain PET imaging.

Speaker: Dr Claude Comtat (Frédéric Joliot Hospital Facility, CEA/DSV/DRM, Orsay, France)

09:45 Voxel-by-voxel analysis of brain SPECT perfusion in fibromyalgia

Fibromyalgia (FM) syndrome is a chronic pain illness characterized by widespread musculoskeletal aches, pain and stiffness, soft tissue tenderness, general fatigue and sleep disturbances, without clinically demonstrable peripheral nociceptive cause. If a psychogenic trouble has been initially postulated, recent fMRI studies clearly demonstrated a global dysfunction of central pain processing: similar pressures in patients and controls resulted in no common cerebral regions of activation and greater effects in patients. We evaluated, in this prospective study, brain perfusion SPECT at rest (without noxious stimulation) in a homogenous group of severe FM patients using 99mTc-ECD. We performed a voxel-based analysis in comparison to a control group, matched for age and gender. In such conditions, we hypothesized that significant perfusion abnormalities could be highlighted, objectifying cerebral processing associated with spontaneous pain in FM patients. Methods. 18 hyperalgesic FM women (48-years old; 25-63 yrs; ACR criteria; VAS for pain: 82 ±4) and 10 healthy women matched for age were enrolled in the study. A voxel-by-voxel analysis was performed using SPM2 (p<0.001c). Results. Voxel-by-voxel perfusion analysis exhibited individual and collective brain SPECT abnormalities, including hyperperfusions of the somatosensory cortex and hypoperfusions of frontal, cingulate, medial temporal and cerebellar cortices. Conclusions. In the present study performed at rest in hyperalgesic patients, we found significant hyperperfusions in regions known to be involved in the sensory dimension of pain and significant hypoperfusions in supposed affective dimension- related areas. As current pharmacological and non-pharmacological therapies act differently on both components of pain, we hypothesize that individual voxel-by- voxel SPECT analysis could be a valuable and easily available tool to adapt individual therapeutic strategy and perform an objective follow-up of pain processing recovery under treatment.

Speaker: Dr Eric Guedj (Service Central de Biophysique et Médecine Nucléaire CHU Timone APHM)

10:00 Influence of tumor density on CT attenuation corrected PET:Phantom studies

The use of external positron emitting sources or single photon emitting sources at energies similar to annihilation photons seems to give acceptable results permitting the quantification of PET data. The attenuation correction based on CT data generates artefacts around metallic prostheses, after use of contrast media and in the pulmonary region near calcifications and on dense lesions. In the present phantom study we investigated the density influence of simulated lesions on attenuation correction based on CT transmission data. A cylindrical phantom of 21 cm diameter and 20 cm height, with two compartments, was used. The lower one with cylinders of 4.0 to 10.0 mm to evaluate the resolution of the system and the upper one with 3 pairs of cylindrical vials of 2.6cm diameter and 4.5 cm height, positioned in a distance of 7 cm from the axis of the cylinder in a way to form an hexagon. The first pair of vials was completed with alcohol solution (d= 0.8 g), the second one with water (d= 1.0 g) and the third one with a KJ solution (d= 1.2g). The phantom’s lower part was filled with water and the upper with water or polystyrene spheres plus water containing 18-FDG. In one vial of each pair equal activities of 18-FDG were added, giving a ratio of 3 with the environment water or water plus polystyrene. An initial acquisition by spiral CT was followed by one bed acquisition of emission and external source transmission data. After reconstruction by using filtered back projection for CT, iterative OSEM algorithms for PET and scaling and segmentation attenuation correction procedures, proposed by the constructors, we measured on transversal slices the activities of the 6 vials by drawing similar ROIs over the vials and one ROI in the centre of the slices to estimate the background activity. On non attenuation corrected (NAC) and CT attenuation corrected (CTAC) images the measured and corrected for attenuation activities are: First tomograph NAC d=0.8: 571 counts d= 1.0 : 480 counts d=1.2: 430 counts CTAC 4966 " 4853 " 7589 " Second " NAC d=0.8: 850 counts d=1.0 : 661 counts d=1.2: 574 counts CTAC 7421 " 7167 " 18343 " Third NAC d=0.8: 700 counts d=1.0 : 620 counts d=1.2: 590 counts CTAC 4900 " 3900 " 3300 " In conclusion the activities on NAC images for all the 3 tomographs are inversely depended on the density of the liquids due to more intense absorption by the denser material. On CT attenuation corrected slices, in the two first tomographs, the estimated activity in denser material of specific activity 1.2g is much higher than the expected due to incorrect attenuation correction procedure. In the third tomograph the CTAC has not the same correctional effect.

Speaker: Prof. Dimitris Maintas (Institute of Isotopic Studies, Athens)

10:30 Coffea break

11:00 Internal targeted radiotherapy : hopes and perspectives for therapeutic applications of nuclear medicine

Internal targeted radiotherapy consists in the in situ irradiation of tumour cells using an appropriate radiolabelled agent. Such an agent should be sufficiently tumour-specific in order to accumulate and be retained for a long time in the tumour, while its elimination from normal tissues should occur quickly. Such an approach has already proved to be efficient for the treatment of thyroid cancer (using iodine 131)and lymphomas (using radiolabelled antibodies), and is now being evaluated for many cancer diseases through several ways of improvement. Research currently focuses on the promotion of new targets, the development of new targeting agents and the adaptation of the pharmacokinetic approach for delivery to tumour cells, the use of new radionuclides, and a better comprehension of the radiobiological effects and dosimetry at the cellular level. Among new targets, several antigens and membrane receptors are of interest, especially if binding of the radiolabelled agent leads to its internalization. Concerning the targeting agents, pretargeting systems allowing better tumor/non tumor ratios must be especially emphasized. Among radionuclides, numerous beta-emitters could be used, and more recently the use of alpha emitters also represents a particularly exciting new area of investigation. Moreover, clinical settings for which internal radiotherapy could be efficient and useful are better defined and, together with the possibilities of associating such an approach with other treatment modalities (ie, chemotherapy, hormonotherapy and more generally biological response modifiers), one can foresee that targeted radiotherapy will soon be considered as a promising new treatment tool for cancer.

Speaker: Prof. Jean-Philippe VUILLEZ (CHU GRENOBLE)

11:30 Impact of 18FDG-PET/CT on biological target volume (BTV) definition for treatment planning for non-small cell lung cancer patients

Purpose: To test the feasibility of FDG based PET/CT data on target volume delineation in radiotherapy treatment planning of NSCLC patients, and impact of these outlined biological target volumes (BTV) for IMRT treatment. Materials and methods: Patient diagnosed with non-operable NSCLC in the right upper lobe had a 3D conformal planning based on CT data with our hypo-fractionated regimen of 52.5 Gy in 15 fractions. Planning was redone with fusion of PET/CT data and 3D CT. Three target volumes were created: necrotic BTV (same as seen in CT), proliferating BTV (based on PET signal to background ratio 1:3) and hypoxic BTV (based on PET signal to background ratio of 1:19, believed to be related to anaerobic glycolisis inefficacy in ATP production). Two IMRT plans were created based on these three BTVs with the intention of giving different doses to each BTV. The first plan (“conservative plan”) delivers 52.5 Gy to the necrotic BTV and 65 Gy to the hypoxic BTV. The second plan (“radical change”) delivers 30 Gy to the necrotic BTV, 52.5 Gy to proliferating BTV and 65 Gy to hypoxic BTV. Results: The use of BTVs in IMRT plan seems attractive because it increases dose to targets considered to need higher doses. It reduces considerably dose to the heart and spinal cord, organs considered to limit dose escalation approaches in NSCLC treatment. However, lower dose to the spinal cord comes at the expense of slight increase in the contra lateral lung dose, still way below V20 limit. Conclusions: The “conservative” IMRT approach can be understood as a PET/CT based concomitant boost to the tumor expressing the highest FDG uptake. The “radical” IMRT planning implies a deviation from the traditional uniform dose target coverage approach, with the intention of achieving a better surrounding tissue sparing and ultimately allowing for dose escalation protocols in NSCLC patients. Several issues should be considered before treating patients using PET/CT based BTVs: tumor motion (patients could be scanned with 4D PET/CT and treated with gated RT), dose calculation accuracy with Monte Carlo based treatment planning algorithms, and specific tumor metabolic activity should be imaged with better radiopharmaceutical markers. We also intend to present four current recommendations for tumor outlining using PET: Qualitative Visual Method (Ciernik), CTV = 2.5 SUV units (Paulino and Johnstone), CTV = 40 % Iso of max Uptake Value (Erdi) and Linear SUV threshold function method (Black).

Speaker: Dr Slobodan Devic (McGill Univeristy)

11:45 Very high resolution PET techniques for small animal and human imaging

Several years ago we presented the idea that PET resolution better than the intrinsic range of the positron at good sensitivity was achievable by placing a high resolution detector—ideally taking the form of a small diameter ring—within the bore of a conventional PET detector ring (J. Nucl. Med. Supp. 2000, 41(5):20P, 2001, 42 (5):55P, 102P). While we examined constructing the inner detector from a number of materials that could potentially support high spatial resolution including cadmium zinc telluride and various scintillators, particularly intriguing was the fact excellent performance appeared achievable by using low proton-number (Z) materials such as silicon for the inner detector. Although nearly all interactions result in a Compton scatter, the Compton-scatter cross-section does not drop quickly with increasing energy. Furthermore, energy resolution is still possible by collecting the scattered photon in the outer ring. Detectors having high propensity for Compton interactions followed by escape of the scattered photon are capable of supporting extremely high spatial resolution that is limited by only the range of the Compton recoil electron. (Of course, photoelectric interactions are usable also). Although detector materials having higher Z are useful for packing more detection efficiency into a given volume, on the basis of the same detection efficiency, higher-Z materials suffer from more multiple interactions, which must be resolved to determine the correct coincidence line-of-response. Various instrument configurations ranging from an intrarectal prostate imaging probe constructed using higher-Z LSO to a device for imaging mice at submillimeter resolution using a low-Z silicon inner detector are under active investigation. For the latter instrument simulation studies have shown that image resolution of ~350microns FWHM is achievable with good sensitivity (~1%) while ~1mm FWHM resolution can be achieved with outstanding sensitivity (9%). Both these figures include the effects of F-18 positron range and acolinearity. Resolution was estimated from images reconstructed using filtered backprojection, which has no intrinsic resolution recovery. Results from such Monte Carlo investigations are encouraging and are presently being validated via experiment. To this end a single-slice proof-of-concept PET instrument was constructed using silicon and BGO detectors. Each 2.2cm x 4.4cm x 1mm silicon detector consisted of a 16 x 32 array of 1.4mm x 1.4mm pads. The silicon detectors were placed edgewise (for detection efficiency) on opposite sides of the 4.4 cm field-of-view and the source was collimated to a 1mm thick slice using thick tungsten plates. Measured efficiency was ~0.7x lower than Monte Carlo predictions which is fully explained by the detectors being biased slightly lower than depletion and by the coincidence timing window. Most encouraging was the spatial resolution which ranged between 700 to 800 microns FWHM across the field-of- view. Even though the high spatial resolution predicted in simulation studies is borne out by experiment, construction of an instrument supporting such resolution at high efficiency remains challenging but not outside the realm of practicality.

Speaker: Neal Clinthorne (University of Michigan)

12:00 Scintigraphic Calf perfusion symmetry after exercise : a new non-invasive index for prediction of cardiovascular events ?

Background. Whole body exercise thallium scintigraphy can detect silent or symptomatic PAD. Whether exercise thallium perfusion muscular asymmetry in the legs has prognostic value is unknown. Methods and Results –Three hundred fifty-eight consecutive patients (mean age 58.8±10.2 year; coronary artery disease, 202/358; 56.4 %) were prospectively followed after thallium 201 myocardial scintigraphy. Scintigraphic calf perfusion symmetry after exercise (SCPSE) was measured at the end of a treadmill exercise test.During the follow-up period (mean, 85.3±32.8 months, range, 6-115), 93 cardiovascular events and cardiovascular deaths (incident cases) occurred. Among incident cases, the percentage of subjects with higher SCPSE values (third tertile) was 45.2 % compared to 29.1 % in controls (p=0.005). In stepwise multivariate analysis performed with the Cox proportional hazards model, previous CAD and SCPSE were the only significant independent predictors of prognosis. The multivariate relative risk of cardiovascular death or cardiovascular event in subjets with higher values of SCPSE was 1.94 (95 % CI: 1.15 - 3.21; p<0.01). Conclusions – This highly reproducible index which is easily and quickly calculated following exercise myocardial scintigraphy could be used as a new tool for identifying high cardiovascular risk patients.

Speaker: Dr Philippe Tellier (CMNA Clinique Sainte Catherine)

12:15 Radioisotope guided surgery with imaging probe, an hand-held High resolution gamma camera

Aim. Radio-isotope guidance is useful in several fields of surgery, such as sentinel node biopsy, parathyroid adenoma and of some neuroendocrine tumours, withdrawal of laterocervical lymph nodes invaded by thyroid cancer. Radioguided surgery (RGS) is generally performed with a gamma probe (GP) providing an acoustic signal proportional to the detected activity. The Imaging Probe (IP) is an hand held, small field of view (FOV) high resolution (HR) gamma camera. Our IP (Li-Tech, Italy) is based on multi-crystal and Hamamatsu position sensitive phototubes (PSPMT) technique with spatial resolution of 2 mm. IP, collimators and crystal-collimator-PSPMT matching are covered by 5 US and EU patents. We used our IP into operator theatre on different fields of surgery to validate the usefulness of HR images in RGS. Patients and Methods. We studied 100 patients with breast cancer for sentinel node detection, 5 patients with parathyroid adenoma, 5 patients operated for thyroid cancer, 5 patients undergoing fine needle aspiration of thyroid notches, 4 patients with neuroendocrine tumours. Devices: State of the art Anger cameras: Philips Skilight and GE Millennium; 1 inch2 and 4 inch2 IP ( Li-thech Italy). Radiopharmaceuticals (RPh): sentinel lymph nodes were detected with 99mTc- nanocolloids, parathyroid adenomas with 99mTc-sestamibi, thyroid nodes and metastases with 123I-Iodine or 99mTc-sestamibi. Neouroendocrine tumours were detected with 111In-octreotide or with research RPh such as 99mTc-Bombesin. Results. IP detected lesions during operations with high sensitivity and specificity in 30 ± 15 sec. this detection time was fully compatible with hand holding of IPs, without mechanical arm help. Sentinel node IP detected multiple nodes in 60% of patients with breast cancer, versus 24% of Anger camera + GP (P<0.01). Nodes detected by IP contained metastases on 30% of patients versus 22% of patients studied with Anger Camera + GP ( ns). Operation time was 7,4 ± 2.8 min with IP, 11.25± 4.25 with GP (P<0.025). IP detected cervical node metastases of thyroid cancer in 3/5 patients: metastases had been missed by Anger camera in one patient. In 5/5 IP correctly detected parathyroid adenomas during operation Anger camera + GP correctly guided the surgeon in 4/5 cases. All thyroid notches were correctly detected and surgeons were correctly addressed for removal of neuroendocrine tumours. Conclusions. HR images and hand held cameras are of valuable help in radioguided surgery. IP is of simple use and significantly decreases the operation time.

Speaker: Prof. Francesco Scopinaro (University "La Sapienza", Rome, Italy)

13:00 → 14:00 PHILIPS Technical presentation :Implémentation du temps de vol dans un TEP CT recherche et clinique

Speaker: Dr York HAEMISCH (Philips)

14:00 → 15:00 Poster session : Imaging systems, Molecular Imaging

14:00 188 RHENIUM-INDUCED CELL DEATH AND APOPTOSIS IN A PANEL OF TUMOR CELL LINES

Rhenium-188 (188Re) represents an interesting radioisotope in cancer therapy for its β- radiation of high energy and γ-ray useful in scintigraphic applications. The genotoxic effect of 188Re-perrhenate has been evaluated on a panel of tumour cell lines (U87,H460,MCF-7,LNCaP,A549,M5076) established from tumors of different type (glioblstoma, lung, breast and prostate cancer) in terms of inhibition of cell viability, micronuclei (MN) and apoptosis. Cells were treated with different activities of 188Re-perrhenate (50-200 μCi) and harvested 48-72 hours later for MTT staining. All the tested cell lines established from lung, breast and prostate cancer seemed to be particularly sensitive to treatment except U87 glioblastoma cells, which behaved as radioresistant. On the contrary, micronuclei induction in binucleated cells was significantly higher in the U87 glioblastoma cells compared to LNCaP and H460 after 48 hours exposure to 25-100 μCi. Induction of apoptosis assessed by apoptotic bodies or TUNEL staining, showed that MTT radioresistance observed in U87 cells correlated well with a very low extent of apoptosis. These results suggest that the evaluation of apoptosis induction rather than induction of MN is a useful tool to asses resistance/sensitivity to treatment with 188Re. In addition, these preliminary results indicate that as a next step radiopharmaceuticals containing 188Re may be successfully used to target lung, breast and prostate tumors in “in vivo” mouse system. ACKNOWLEDGEMENTS Supported by Gruppo V, INFN, Italy.

Speaker: Dr Andrea Perrotta (INFN,Dipartimento di Fisica, Bologna)

14:00 A Detector for Radionuclide Bronchial Scanning

This paper is aimed to describe a sub-miniature scintillation detector suitable for functional inspection of bronchi inside a bronchoscope. The device is made by a probe having diameter adequate for insertion in a bronchial catheter and an external electronics for signal’s analysis. The probe is composed of a scintillation crystal for high-efficiency photon detection and a Si-avalanche photodiode for high quantum-efficiency scintillation readout. The device can be adequately shielded to select a field of view reducing the background and operates in single-photon mode giving the possibility of good energy discrimination. The localization technique takes mainly advantage (i) from the inverse dependence of crystal efficiency on the square of crystal-to-radiation source distance, (ii) the lesion-to-background uptake ratio and (iii) from the exponential absorption of radiation coming from distant locations, eventually reduced by additional selective shielding. Furthermore, the pulmonary environment is characterized, with respect to the surrounding organs, by a reduced specific density of tissues (about 0.7 g cm-3 at maximum expiration and about 0.3 g cm-3 at maximum inspiration). This produces an advantage in terms of lesion visibility due to high lesion-to-background ratio. A somatostatin-analogue depreotide, labelled with 99mTc, is currently used for detection and characterization of pulmonary nodule that means 140keV single photons have to be detected. Previous measurements concerning a laboratory setup made by a 57Co point-source and a 2.85x2.85x5.00mm3 CsI:Tl scintillator have been made at different distances. Furthermore, the limits of the technique have been assessed by Monte Carlo simulations. The considered geometry, reproducing a very severe clinical situation, consisted of a duct positioned at the centre of a homogeneous cylindrical background-source (40cm diameter, 30cm height) adjacent to a homogeneous lesion- source 0.2cm diameter. Results gave a very small lesion to background factor which has to be enhanced 103 times to obtain a good detectability. New Monte Carlo cases are presented concerning pulmonary configurations and nodule diameter in the range from 1cm to 4cm. Si-avalanche photodiodes with reduced peripheral dead zones customized by Hamamatsu are used for new measurements with phantoms simulating clinical conditions. Experimental results are compared with simulated data.

Speaker: Dr Raffaele Scafè (Casaccia Research Center, ENEA, Rome, Italy)

14:00 A LSO Beta Microprobe for Measuring Input Functions for Quantitative Small Animal PET

A miniature scintillation microprobe has been developed to measure the input function in live rodents for use in longitudinal, quantitative PET studies. The probe consists of a small lutetium oxyorthosilicate (LSO) crystal measuring typically 0.3-0.5 mm diameter x 0.5-2 mm in length that is used to directly detect positrons in the blood or tissue. The probe has a sensitivity of 10-30 Hz/mCi/cc and is primarily sensitive to short range positrons emitted by labeled radiotracers in the blood. The sensitivity to gamma-ray background can be minimized using a variable threshold in the readout to discriminate between positrons and gammas. The probe was implanted in one of the tail veins of a Sprague-Dawley rat and the input function was measured for the injection of 0.8 mCi of FDG in the other tail vein. The probe exhibits a fast time response that is able to quickly and accurately measure the concentration of 18F circulating in the bloodstream. Additional tests were also carried out to study the probe’s sensitivity to gamma ray background.

Speaker: Dr Craig Woody (Brookhaven National Lab)

14:00 A new multimodal and quantitative approach for in vivo small animal brain studies : combination of the Nuclear Magnetic Resonance and the radiosensitive Beta-MicroProbe

Elucidating physiological mechanisms in small animal in vivo requires the development of new techniques including imaging with multiple modalities. Combining exploratory techniques has the tremendous advantage to record simultaneously complementary parameters on the same living animal. In this field, an exciting challenge remains in the combination of Nuclear Magnetic Resonance (RMN) and Positron Emission Tomography (PET) since small animals studies are limited by strict technical constraints in vivo. Therefore, we proposed to couple NMR techniques to the radiosensitive Beta-MicroProbe, that showed to be an elegant and versatile alternative to PET measurements. This coupling offers the possibility to obtain complementary and simultaneous biological informations and to realise an absolute quantification of the measured radioactive signal in a brain structure of interest. Previous works allowed us to validate the physical feasibility of this combination. Indeed, we demonstrated experimentally and by Monte Carlo simulations that, although the detection volume of the probe could be significantly reduced because of the influence of the intense magnetic field on the positrons range, the sensitivity is only slightly reduced (<7% for a 500 µm diameter probe in 18F). In fact, two effects have been pointed out. First, the magnetic field improves the detection of the local efficiency for the positrons close to the probe, compensating partially the non-detection of the positrons far from the probe. Second, an increase in the deposited energy in the scintillating fiber leads to an increase of the light output. Based on our physical results, we evaluated the feasibility of this new dual- modality system in a biological context. To this aim, we decided to reproduce pharmacological measurements with the probe on anesthetized rat with [18F]-MPPF, an antagonist of the 5-HT1A receptors, but with animals placed during the acquisition inside the 7-T magnet. Two probes were chronically implanted for each rat, one in the hippocampus (specific area) and one in the cerebellum (nonspecific area). We attested the stability of [18F]-MPPF kinetics on both brain areas compared to experiments achieved without magnetic field. We thus demonstrated the possibility to realise a tracer biokinetic measurement simultaneously to the acquisition of anatomical images of the hippocampus and cerebellum. Finally, this study raised a question of quantification particularly interesting since the detection volume of the probe is not entirely included in the hippocampus. Therefore, using a voxelised rat brain phantom and the precise knowledge of the probes positions attested with aMRI images, the present experiments were also conducted to improve the absolute quantification of the tracer concentration in the hippocampus.

Speaker: Ms Aurélie Desbrée (Laboratoire Imagerie et Modélisation en Neurobiologie et Cancérologie)

14:00 A simple method to estimate coordinate resolution of pixelized detectors

A simple method to estimate the coordinate resolution of pixelized detectors using non-single photon counting read out is proposed. This method can be useful particularly for testing the detectors used in digital imaging systems. The traditional technique of determining the coordinate resolution of this type of detectors is mainly based on the use of line spread function measurements that permit to recover the resolution distribution. The proposed method is based on the determination of the statistical correlation between neighbor elements in the detector. The correlation is determined by means of statistical noise measurement of the isolated elements and the linear combination of neighbor elements. To suppress the possible contribution of the beam spatial variation, the differences between neighbor elements is used. The distribution of differences between the detector neighbor pixels is constructed for the pixels sharing the same edge “X” and “Y”, or having only one common vertex “U” and “V”. For the validation of the method, the pixel detector of mammography unit General Electric Senographe 2000D has been used. For this purpose the pixelized detector has been irradiated by X-ray photons along the entire surface. The dependence of the resolution on the photon intensity value and the pixel coordinate has been studied. The asymmetry of correlation between the perpendicular directions has been studied to estimate the possibilities of this method to detect the existence of asymmetry in hardware design. The coordinate resolution of the studied detector, having pixel size of 100 µm, is of the order of 50 –55 µm, that is almost twice larger than the value conditioned by pixel geometrical sizes. A small asymmetry between the X and Y directions has been detected. By use of detector spatial resolution data in the approximation of Gaussian behavior of point spread function the modulation transfer function can be estimated. The reconstructed modulation transfer function is in good agreement with MTF measurements of the studied detector.

Speaker: Dr Varlen Grabski (Instituto de Fisica UNAM)

14:00 A YAP-CAMERA FOR MONITORING THE BIODISTRIBUTION OF RADIOTRACERS CONJUGATED WITH HYALURONIC ACID IN “IN VIVO “SYSTEMS

The aim of the SCINTIRAD experiment is to determine the radio-response to 188Re in various cells in vitro, the biodistribution of the radiopharmaceutical in different organs of mice in vivo, and subsequently the therapeutic effect on liver tumours induced in mice inoculated with M5076 murine fibrosarcoma cells. Both the gamma and beta emissions of 188Re have been used in the experiment. The in-vivo biodistribution in mice is studied with a gamma camera using different parallel hole collimators. In the 188Re spectrum, while the 155 keV gamma peak is useful for imaging, less intense photon lines emitted at larger energies (given their much larger penetration) and the beta particles emitted in the decay (given the secondary interactions they can generate everywhere in the FOV) act as noise from the point of view of the image reconstruction. Gamma cameras previously used to image biodistributions obtained with 99Tc are therefore not optimized for use with 188Re. A new setup of the gamma camera has been studied for 188Re: 66x66 YAP:Ce crystals (0.6x0.6 mm2, 5μm optical insulation between crystals, thickness 10 mm) guarantee a FOV of 40x40 mm2; a Hamamatsu R2486 PSPMT, 3 inch diameter, converts their light output into an electrical signal and allows reconstructing the spatial coordinates of the light spot; the direction of the incoming photon is selected by a Lead collimator with 1.5 mm diameter parallel hexagonal holes, 0.18 mm septa, and 40 mm thickness. Using the new setup, preliminary results have been obtained both with 99Tc filled capillaries and 188Re filled capillaries and small phantoms. The energy spectrum of the collected photons and some preliminary measurement of the spatial resolutions achievable with the two sources will be presented. In addition the first experiment of imaging with a mouse in vivo is planned well ahead the beginning of the conference and will be reported. The preliminary results obtained so far suggest that radiopharmaceuticals containing 188Re and HA may represent useful tools for the treatment and imaging of liver tumours. Supported by INFN, Gruppo V.

Speakers: Dr Andrea Perrotta (Dip. Fisica e INFN, Bologna), Prof. Francesco Navarria (Dip. Fisica e INFN, Bologna)

14:00 ADVANCED HIGH-SPEED MODULAR SYSTEMS WITH NEW INTERCONNECTS

Development of Electronics, Communication and Computer technologies are moving to full integration of these fields into compact modular systems, integrated by modern networks. High-frequency processor are not more a basic way for advanced high- performance systems and multi core processors become more perspective, and tradition parallel bus based system architectures (VMEbus, cPCI/PXI) are not more perspective for new generation high-speed systems and some new serial approaches to advanced high-performance modular systems are required now. Today existing resources of tradition modular systems should be used in compatible new generation modular systems as much as possible, including mechanics and electronics in first order. Fundamentals in system architecture development are compact component approach, low power processor with new serial high-speed interface chips and high-modular structure on all levels of both system and network interconnects. There are many parameters that define advanced system design, including basic microprocessor type and interfaces, network architecture and topology, basic interconnects and node interactions modes in distributed system. Advanced modular system components and development approaches based on new international standards are described and discussed, including new high-speed serial interconnects, module structures, new connectors, interactions of processor cores in distributed compact nodes and general network architecture and topology. Perspective tendency is especially effective for high-performance data acquisition and image processing system development in any application fields.

Speaker: Dr Vyacheslav Vinogradov (Institute for Nuclear Reserarch RAS)

14:00 An Intra-operative Compact Gamma Imager for Radio-guided Cancer Surgery : Evaluation for Sentinel Node Localization

In addition to intra-operative probes, compact gamma cameras are very attractive to provide more efficient radio-labeled tumors localization. In this context, our group has developed a high resolution intra-operative imager, POCI (Per-Operative Compact Imager) dedicated to nuclear medicine purposes. The current POCI prototype is based on a 40 mm diameter intensified position sensitive diode coupled to a gamma head module. The imaging performances at 140 keV gamma energy are a spatial resolution of 2.3 mm FWHM and a corresponding detection efficiency of 10.7 cps/µCi. This camera is now the subject of a hospital program for a clinical research (clinical protocol N° P040417) including 200 patients in collaboration with the Tenon hospital (Paris). POCI is currently evaluated in sentinel node detection protocol for breast cancers staging. Preliminary results, obtained both in nuclear medicine department and theater block, already demonstrate that the imaging performances of POCI suit intra-operative imaging conditions as far as compactness, exposure time and spatial resolution are concerned. Overall imaging performances and first clinical evaluations will be extensively presented and discussed.

Speaker: Dr Stephanie PITRE (IMNC, Orsay, France)

14:00 An Intraoperative beta probe for brain tumor surgery

Surgery is still considered the primary therapeutic procedure for high grade gliomas and several recent clinical studies have shown that gross total tumor resection is directly associated with longer and better survival when compared to subtotal resection. In order to refine the resection especially in the boundaries of gliomas, we are developing an intraoperative probe specifically dedicated to the localization of residual tumor labeled with positron emitters. The probe was designed to be compact and electrically safe in order to be directly coupled to the excision tool leading to simultaneous detection and removal of tumor tissues. The detection head of the intraoperative beta-probe consists of plastic scintillating fibers held in a close packed annular arrangement ensheathing the excision tool. This head is connected to an optic fiber bundle that exports the scintillating light to a multi-channel photomultiplier. The annihilation gamma ray background generated by the annihilation of beta+ in tissues is eliminated by a real-time subtraction method. Validation of the technical choice and optimization of the probe geometry were performed by preliminary measurements and Monte Carlo simulations. Simulations were realized using MCNP code and an anthropomorphic brain phantom filled with different radiotracer activities. Optimal performances were obtained with a detection head composed of 2-mm diameter and 0.5-mm long scintillating beta-sensitive fibers associated to 2-mm diameter and 2-mm long beta-shielded fibers for the gamma ray background rejection. The theoretical probe sensitivity was found to be 1.95 cps/kBq/ml with a gamma ray rejection efficiency of 99.6%. The expected minimum radiotracer detectable concentration for 18F-FET was 3.7 kBq/ml. When compared to the 10.7 kBq/ml average concentration in the bulk of the tumor, this value demonstrate the ability of the probe to define more accurately the extent of brain tumor resection. Following these results, a first prototype of the probe based on seven detection elements was developed. Characterization of the device including uniformity of the signal gain and of the beta and gamma sensitivity among the detection elements, spatial resolution and gamma ray rejection efficiency will be presented. The ability of the prototype to localize small amounts of beta+-emitting radiopharmaceuticals will be also demonstrated using a brain phantom simulating a surgical cavity after excision of the bulk of the tumor labeled with 18F.

Speaker: Dr Laurent MENARD (Laboratoire Imagerie et Modélisation en Neurobiologie et cancérologie (UMR 8165) - Université Paris 7)

14:00 ANALYSIS OF DIFFERENT DETECTOR AND ELECTRONICS DEFECTS ON F18-FDG IMAGES

Objectives Aim of the present work is to describe and quantitatively assess on reconstructed FDG-PET images, the effects of different levels of detector failures for the HR+, a full ring block designed scanner. Methods All the studies were performed in 2D mode (span 15) with the ECAT EXACT HR+ PET scanner (CTI-Siemens, Knoxville-USA). By using data acquired with PET scanner fully calibrated and perfectly functioning different detector defects were simulated: 1) one block detector with a 40% (B1 40%) and 100% (B1 100%) loss of efficiency; 2) one analog board involving 2 contiguous axial blocks, with 40% (AB 40%) and 100% (AB100%) loss of efficiency; 3) two opposite blocks in the same plane with 40% and 100% loss of efficiency (B1 40%B2 100%) and (B1 100%B2 100%). Different radioactivity distributions were also considered: -the 20 cm diameter uniform phantom containing a solid solution of 68Ge; scanned at high counting statistics and reconstructed with FBP ramp filter, -the anthropomorphic Alderson phantom filled with different 18F radioactive concentration and containing 4 spherical lesions acquired at clinical counting statistics and reconstructed with OSEM 2 iterations and 8 subsets (2i/8s) as routinely used in whole-body PET FDG studies and (4i/16s). The analysis of the faulty effects was performed taking as reference the original images without defects. Images were evaluated qualitatively, by visual inspection of expert observers, and quantitatively by drawing ROIs to assess % difference of SUV (ΔSUV%.) Results Visually and quantitatively, in spite of evident defect on sinograms (cold diagonal bands), reconstructed images are affected by slight artefacts when 1 block or 2 axial contiguous blocks even with 100% efficiency loss is considered. Artefacts appear as cold stripes following the Fan angle covered by the defective detector. More important variations are observed when two opposite blocks in the same plane, both with 100% inefficiency, are considered. In this case, depending on the geometrical position of the two blocks, a well structured artefact along the direct coincidence line is visible and a peak-valley count distribution affects the image. To this case |ΔSUV% | maximum values of 5.4% and 15.1% are respectively measured with 2i/8s and 4i/16s OSEM. Conclusions This study showed that for HR+ scanner, even a 100% loss of efficiency of a single block or an analog board (2 axial blocks) particular evident on sinograms does not substantially affect qualitatively and quantitatively FDG whole-body reconstructed images (SUV % variations less than 4%). When two opposite blocks in the same plane have 100% loss of efficiency larger inaccuracy should be expected and in this case stopping clinical activity is claimed. However, all defective block considerations referred as to be not relevant for scanner clinical use must be utilised to avoid practical problems in temporary patient management and not to change maintenance scheduler or lower the quality standards. Effects of malfunctioning blocks and electronics on 3D acquisition images is under evaluation.

Speaker: Dr Felicia Zito (Nuclear Medicine Department, Fondazione Ospedale Maggiore Policlinico, Mangiagalli e Regina Elena, Milan, Italy)

14:00 Assessment of the impact of x-ray tube voltage on quantitative analysis of neurological PET when using CT-based attenuation correction

The advent of dual-modality PET/CT imaging had great impact on improving the value of diagnostic PET in localizing, evaluating and therapeutic monitoring of head and neck cancer and equally valuable for other localizations that are difficult to pinpoint. In addition, the use of CT images for CT-based attenuation correction (CTAC) of PET data decreases overall scanning time and creates a noise-free attenuation map (μmap). The most common CTAC procedure requires a bi-linear calibration curve acquired under standard imaging conditions to convert the patient’s CT image from low effective CT energy into an attenuation map at 511 keV. Given that different tube voltages are used in current PET/CT scanning protocols depending on patients size and the region under study, this work was designed to provide answers to the legitimate question of the clinician or physicist: “what is the magnitude of error due to acquiring CT at e.g. 80 kVp when the calibration curve would be the manufacturer’s standard of 120 or 140 kVp?” and vice versa, that is, acquiring CT images at specific tube voltages and varying the voltage for derivation of calibration curves. The impact of using a single calibration curve on the accuracy of CTAC for images acquired at different tube voltages was investigated through quantitative analysis of created μmaps, generated attenuation correction factors and reconstructed neurological PET emission data using experimental anthropomorphic phantom and clinical studies. For CT images acquired at 80 and 140 kVp, an average relative difference of -2.9% and 0.7% with the images acquired at 120 kVp, respectively, was observed between absolute activity concentrations in five regions of the anthropomorphic striatal phantom when CT images are converted using a single calibration curve derived at 120 kVp. Likewise, an average relative difference of 1.9% and -0.6% was observed when CT images are acquired at 120 kVp and CTAC uses calibration curves derived at 80 and 140 kVp, respectively. It was concluded that using a single calibration curve derived under standard scanning conditions during the CTAC procedure to images acquired at different tube voltages does not affect significantly the visual qualitative interpretation and quantitative analysis of neurological PET emission images. The same behaviour was observed when calibration curves are derived at different tube voltages and used for conversion of CT images acquired at fixed tube voltage. These results might contribute to alleviate the quality assurance procedures required for daily operation of PET/CT scanners in a clinical environment.

Speaker: Dr mohammad reza Ay (Department of Medical Physics, Tehran University of Medical Sciences, Tehran, Iran and Division of Nuclear Medicine, Geneva University Hospital, Geneva, Switzerland)

14:00 Beam hodoscope for online PET in hadrontherapy

The feasibility of a PET prototype for online and in-situ dose deposition control in hadrontherapy is under study within our group at the “Institut de Physique Nucléaire de Lyon”: IPNL. This is a non conventional use of the PET techniques that would allow a real time non invasive control of the execution of a given cancer treatment plan with heavy charged ionizing particles beam. In practice, our experimental setup is a somehow conventional PET scanner equipped with a transparent and fast beam hodoscope detector. Time correlations between the arrivals of beam particles to the target and prompt nuclear gamma rays detection as well as the annihilation photon detection from produced β+ emitter isotopes are exploited for an online and real time 3-D reconstruction. We will report on the measurement performed with our prototype beam hodoscope that has the ability of : • high counting rate , up to 10^8 particles per second • high transversal coordinates measurements, 250 μm • high time stamp resolution, below 1ns

Speaker: Mr Anthony BUTHOD (Institut de Physique Nucléaire de Lyon ( Groupe Imagerie bio-médicale ))

14:00 CdZnTe Detectors for the Positron Emission Tomographic Imaging of Small Animals

We propose here is to study the suitability of a new solid state detector like cadmium telluride (CdTe and CdZnTe) detectors to address the spatial resolution issue associated with small animal Positron Emission Tomography (PET) systems. The main advantage of semi-conducting detectors over scintillation detectors lies in the fact that the segmentation can be obtained very easily thanks to the electrode pixellisation as small as desired. We have jointly developed a specific three-dimensional CdZnTe (CZT) detector geometry (16x20x0.9 mm3 CZT detectors equipped with specific orthogonal strips including 16 anodes and 5 cathodes with 1 and 4 mm pitch respectively) for a transverse irradiation and a preamplifier stage to achieve the best coincidence timing performance between two CZT detectors. The encouraging CZT-CZT coincidence time of 2.6ns FWHM obtained with planar detector has been confirmed with orthogonal strips CZT-BaF2 coincidence measurements. In a second step, we simulate the spatial resolution and detection efficiency performance of a realistic stacked CZT detector module with depth of interaction (DOI) capability. Preliminary simulations indicate that the proposed design could outperform an LSO-based system, with a better homogeneity of the spatial resolution across the Field Of View (<1 mm FWHM up to 44 mm off the FOV center). The efficiency obtained for parallelepiped detectors, 40mm CZT equivalent to 10mm LSO, could be greatly improved, especially pertaining to the homogeneity across the field of view, if a new trapezoidal geometry for the CZT detector could be implemented. With the objective to confirm the simulation results, a multi-channel experimental bench has been set up. It consists of two detectors put one behind the other to obtain a total depth of 40mm. The 16 anodes and 10 cathodes are connected to 26 identical preamplifiers. The generated signals are fed into customized constant fraction discriminators put on 8 electronic boards, each controlled by an FPGA. The data are processed through a Labview program. CZT-BaF2 coincidence times of 2.1ns FWHM on anodes and 1.6ns FWHM on cathodes have been obtained. After energy calibration and fine tune of threshold, the spatial distribution appears homogeneous along cathode directions and follow the law of absorption along anode directions. Work is in progress now to confirm experimentally the good spatial resolution homogeneity due to the DOI and detection efficiency obtained by simulation. Simulations at the scale of a whole small animal PET system coupled to an experimental validation of these results are under way. This new 3D CZT detector geometry may open up new vistas for innovative system architectures, particularly regarding the global sensitivity improvement.

Speaker: Dr LOICK VERGER (CEA-LETI)

14:00 Characterisation of a high spatial resolution SPECT dedicated to small animal imaging

The interest of using small animal imaging has been primarily due to recent advances in genetics with the need to perform longitudinal in vivo studies on mice or rats. Recent works have shown that Single Photon Emission Computed Tomography (SPECT) imaging of small animals can provide high spatial resolution. In our institute, we are developing A Multi modality Imaging System for Small Animal (AMISSA) combining X-ray, SPECT and PET devices. In this paper, we focus on the SPECT component. It is based on four cameras arranged around the animal in ring geometry. Each camera consists of 5 independent detector modules arranged on a ¼-circle pointed to a 0.5mm pinhole aperture. Each module is built with a 8x8 scintillator arrays of 2.3x2.3x28mm³ YAP:Ce pixels optically isolated and glued to a 8x8 multianode PMT connected to a dedicated electronics. The data are sent to a PC through a USB port. The spatial resolution obtained after a dedicated cone beam reconstruction is 1.3mm³. The system energy resolution is 30% at 140keV. The performances and the first experimental results will be presented.

Speaker: Mr Virgile Bekaert (IPHC/CNRS - ULP)

14:00 ClearPET, a high performance small animal PET scanner

The ClearPET is a high performance small animal PET scanner that has been developed within the Crystal Clear Collaboration (CCC) and is now commercially available to customers worldwide through the Raytest group (http://www.raytest.com). Its high sensitivity and spatial resolution are achieved thanks to a patented phoswich configuration made of two types of scintillating crystals: L(Y)SO and LuYAP:Ce, thus providing Depth-of-interaction (DOI) information. In the commercial version 8x8 phoswich units, each one composed of one L(Y)SO crystal coupled to one LuYAP (both of dimensions 2x2x10 mm3), are glued to multichannel Photomultiplier tubes (Hamamatsu R7600). A unit of four PMTs arranged in-line represents one of 20 sectors of the ring design and is designed as a “cassette”. The gantry on which the 20 cassettes are fixed allows for rotation of the detector modules around the field of view. Free-running ADCs digitize the PMT pulses and later digital data processing determines the gamma energy, the phoswich layer (DOI) and the exact pulse starting time, which is subsequently used for coincidence detection. The opening diameter of the ring can be adjusted to 140 or 260 mm. In the first case, the animal port is 125 mm, what is appropriate for whole body rodent studies whereas the second case allows for an animal port of 245 mm suitable for primate brain studies. The field of view axial length is 110 mm. In order to illustrate the performances of the commercial ClearPET, the first results obtained (spatial resolution, sensitivity, count rates and phantoms images) will be presented in this poster as well as the updated situation of the first commercial unit sold to the Animage platform in Lyon.

Speaker: Dr Pablo Sempere Roldan (Raytest France)

14:00 Clinical evaluation of pixelated NaI:Tl and continuous LaBr3:Ce compact scintillation cameras for breast tumors imaging

Although scintimammography was introduced more than 10 years ago, it has never become routine in the majority of Nuclear Medicine Centers. The principal limiting factor in the clinical acceptance is certainly its low sensitivity for cancers sized < 1cm, mainly due to the lack of equipment specifically designed for breast imaging. The very low sensitivity of scintimammography for tumors under 1 cm diameter is not trivial, because the ability in visualizing small breast cancers is really crucial for the future development and clinical acceptance of scintimammography. To this aim a number of dedicated gamma cameras with superior imaging performances were specifically designed for the breast. Hundreds of clinical trials have been currently performed by dedicated gamma cameras based on Position Sensitive Photo-Multiplier Tubes (PSPMT) coupled to scintillation arrays and cadmium-zinc-telluride (CZT) semiconductor cameras. In particular the latter shows higher performances due to 6.5% of energy resolution, that causes a better scattered radiation rejection. Finally INFN has been developing a new scintillation camera based on the latest scintillator generation, LaBr3:Ce, that is demonstrating superior imaging performances than CZT detector with comparable energy resolution. In order to foresee future advances in functional breast imaging, in this paper we propose a clinical comparison and evaluation between this gamma camera with the one previously developed under “IMI” Italian project for technological transfer of INFN. The gamma camera, made by CAEN and Pol.Hi.Tech, has an overall dimension of 112x120x75.3mm3 and consists of an array of 42 one in. PSPMTs Hamamatsu H8520-C12 closely packed, a NaI(Tl) scintillation array (1.8 x1.8x 6mm3 pixel) and a general purpose collimator. A clinical experience is performing by this gamma camera on a number of patients with breast cancer suspicion.. Latest generation detector consists of continuous LaBr3:Ce scintillator coupled to a Hamamatsu H8500 Flat Panel PMT. The planar LaBr3:Ce integral assembly, realized by Saint Gobain, is 50.8×50.8 mm2 size and 4 mm thick with a 3 mm thick glass window. Basic detector principle is the same of Anger camera with scaled dimensions. The detector showed 1 mm intrinsic spatial resolution and 70% detection efficiency. In this paper we show how high spatial resolution, high energy resolution and continuous position response improve detector imaging performances to better categorize the lesions by the morphology of the radiotracer distribution in the breast tissue. The clinical experiment consisted of the co-registration of the breast examination from the two dedicated cameras. The comparison between standard and high resolution images of two cases of breast cancer suspicion, highlights how Anger camera poorly imaged malignant lesions. Otherwise, the high resolution scans produced by LaBr3:Ce camera showed higher tumor contrast with a detailed imaging of uptake area than pixelated NaI(Tl) dedicated camera. Furthermore a dramatic increase of Signal to Noise Ratio (SNR) of lesion with a consequent strong improvement of tumor detectability. represents the most impressive result, bearing in mind that the definition of a breast lesion malignancy is determined by a focal concentration of radioactivity (hot spot) in the breast tissue. These results were confirmed by byoptical findings.

Speaker: Prof. Roberto Pani (INFN Dept Experimental Medicine and Pathology - University of Rome "La Sapienza"-Italy)

14:00 Comparing assessments of myocardial viability between PET and MRI

There are an increasing number of patients with disabling heart conditions related to left ventricular dysfunction, and two thirds of these cases are the result of coronary artery disease. A reduction of mortality and morbidity may be achieved if these diseases can be diagnosed correctly and treated at an early stage before symptoms occur. It is well established that impaired left ventricular (LV) function does not necessarily represent irreversible tissue injury, because contractile performance can improve after revascularization. Correct assessment of the extent of viable and nonviable myocardium in patients with severely reduced LV ejection fraction (EF) and chronic coronary artery disease is important for clinical decision making, because perioperative mortality and morbidity of these patients are increased. However, it has been shown that patients with dysfunctional but viable myocardium profit most from revascularization. Detection of viability based on assessment of myocardial perfusion and glucose metabolism by PET is presently considered by many study reports. The extent of viable myocardium correlates with improved contractile performance after revascularization and affects both short-term and long-term prognosis. MRI, using gadolinium-based contrast agents, delineates irreversibly damaged myocardium and predicts areas that will not recover functionally after revascularisation. Accurate diagnostic information to identify viable myocardium in patients with LV dysfunction who have viable myocardium is the cornerstone to decrease the mortality and morbidity of these patients. Cardiac positron emission tomography and cardiac magnetic resonance imaging are two powerful tools for predicting which patients will have an improved outcome from revascularisation. The aim of this study was to integrate the functional and anatomical information get from PET and MRI in a simulation study, respectively, and could provide diagnostic information clinically by this study in expectation.

Speaker: Ms Ching-Ching Yang (Institute of Radiological Sciences, National Yang-Ming University)

14:00 Comparison of Monte Carlo simulated and measured performance parameters of miniPET scanner

In-vivo imaging of small laboratory animals is a valuable tool in the development of new drugs. For this purpose miniPET, an easy to scale modular PET camera, has been developed at our institutes. The detector module comprises Hamamatsu R8520 position sensitive PMT and LSO crystal matrix (8 x 8 array of 2 x 2 x 10 mm size). The current system has 4 modules, which makes it possible to rotate the whole detector system around the axis of the field of view. Detector signals are processed by digital signal processing and programmed FPGA boards. Data collection and image reconstruction are performed using a data acquisition (DAQ) module with Ethernet communication facility and a computer cluster of commercial PCs. In order to allow flexibility and easy management of various configurations, coincidence conditions were analyzed using digitized detector signals. This signal processing philosophy allowed us to completely avoid any hardware coincidence circuits. The architecture of DAQ system software of the miniPET is like a client- server network, which makes it easy to scale up the instrument and ensures flexibility in building scanners of different diameter and axial field of view (FOV) using identical signal processing boards and DAQ software. Performance tests were carried out to determine system parameters, such as energy resolution, spatial resolution, sensitivity and noise equivalent count rate. Energy resolution proved to be 19.6 % on average while spatial resolution in radial direction ranged between 1.5 and 2.3 mm. Using a small volume source the sensitivity of the miniPET was found to be 1.08 cps/kBq. GEANT4 based GATE Monte Carlo software package was used to simulate PET data analogous to those of the performance measurements. GATE was run on a Linux cluster of 10 processors (64 bit, Xeon with 3.0 GHz) and controlled by a SUN grid engine. The application of this special computer cluster reduced the time necessary for the simulations by an order of magnitude. The simulated energy spectra, maximum rate of true coincidences and sensitivity of the camera were in good agreement with the measured parameters.

Speaker: Mr Sandor Attila Kis (PET Center, University of Debrecen, Debrecen, Hungary)

14:00 COMPUTIS : a new european project.

Full Title : Molecular Imaging in Tissue and Cells by Computer-assisted Innovative Multimode Mass Spectrometry This project aims to develop new and improved technologies for Molecular Imaging Mass Spectrometry (MIMS), enabling innovative approaches in functional genomics, proteomics and metabolomics, as well as for investigation of cells and tissues. Three objectives are considered as part of this project: - Innovate MS imaging instrumentation through the development and application of novel desorption, ionisation and detection techniques - Advanced diagnostic methods for identifying diseases by study of molecular images, - Monitoring of therapeutic effects on expression patterns of damaged and abnormal cells or tissues, This project is based on recent significant improvements in desorption and ionisation techniques such as SIMS (Secondary Ion Mass Spectrometry) and MALDI (Matrix Assisted Laser Desorption Ionisation) associated with Time of Flight mass spectrometers which offer an outstanding ability to analyze organic molecules as large as peptides or proteins at femtomole sample amounts. These techniques are extrapolated to produce actual molecular images of flat specimen by mass spectrometry with a micrometric lateral resolution. This project will provide innovative analytical capabilities for mapping a variety of biological compounds directly at the tissue or cell level by superpositioning information from different sources. Thus MIMS is an extraordinary new tool which could lead to completely new analysis concepts. In order to make it routinely accessible to users, it requires appropriate instrumentation, sample preparation methodology and computerization with high performance massive data acquisition and processing. Application-driven developments of MIMS will be validated as part of this project by close collaboration of pathologists, biologists, analytical chemists and informaticians, to achieve the project objectives and lead to rapid industrial and clinical exploitation owing to the active contribution of the SME's and the industrial partners of the consortium.

Speaker: Dr Serge HAAN (CEA)

14:00 Dual SPECT/MR imaging in small animal

Small animal imaging techniques provides the possibility to follow non invasively and in vivo the morphological or functional evolution of a particular organ or tissue over time. The aim of dual imaging is to combine morphological and functional information. Among multimodality imaging techniques for co-registration of mutual information, SPECT/MRI has not, to our knowledge, be applied practically due to complex interactions between magnetic field and nuclear instrumentation. We describe and illustrate our practical experience in small animal sequential SPECT/MR images acquisition and fusion. For SPECT imaging we used a small animal dedicated rotating single head gamma camera with a 1.5mm diameter tungsten pinhole collimator and 12cm in focal distance (Gaede, Freiburg, Germany) and an algebraic reconstruction method able to achieve isotropic millimeter 1x1x1mm3 spatial resolution [1, 2]. MRI was performed on a low field 0.1T open system (Bouhnik SAS, Vélizy-Villacoublay, France) with an homogeneous (10-6) imaging volume of 10x10x6cm3 [3]. 3D MR sequences are developed on a SMIS (Guilford, UK) console to achieve isotropic resolution of 0.4x0.4x0.4mm3, or high in plane resolution of 0.25x0.25mm² and 1mm slice thickness, in less than 2h acquisition times. As sizes of our dual imaging devices (SPECT 103x80x140cm, low field MRI 72x40x118cm) fits in a single room and as the 5 Gauss line is located at a few centimeters from the magnet, pinhole SPECT camera and anesthesia set were located directly next to the MR magnet, without interfering. Small animal (mouse or rat) is maintained under gaseous anesthesia (isoflurane 0.5-1.5%, 0.3L.min-1) in a warmed-up and non-magnetic technical cell dedicated to small animal imaging (CTI, Minerve, Esternay, France) and adapted for both SPECT and MRI devices. This imaging chamber isolates the animal during its transport and imaging thanks to a bite bar warranting the animal positioning from one imaging experiment to the other. After sequential SPECT and MRI acquisitions, co-registration of images is obtained with the AMIDE [4] software (UCLA, California) helped by location phantoms placed on the cell bed. Advantage of the cell concerns transport of the anesthetized animal from one imager to another minimizing therefore the delay between acquisitions without requiring any more manipulation. During the dual imaging experiments in mice, we were able to keep the animals under anesthesia inside the cell up to 10h without consequences. As a proof of experience of this dual imaging modality, we performed sequential SPECT/MRI of implanted glioma brain tumors in adult swiss nude mice (Transgene S.A., Strasbourg, France). Intra-peritoneal administration of 1mL of 0.5mmol of Gadolinium (Dotarem, Guerbet, France) was followed by IV injection of 0.2mL of 700MBq of 99mTc-Sestamibi (BMS, Rueil-Malmaison, France) before dual imaging. For SPECT, 48 projections of 1min in a 64x64 format and 4.5cm radius of rotation was used. For MRI, 3D T1 and T2 weighted images were acquired, with the high in plane resolution specified previously, in less than 4h. Co-registration of brain SPECT/MR images was then applied after dual modality image reconstruction allowing for differentiation of brain edema and tumor metabolism. [1] Israel-Jost V, Choquet P, Salmon S, Blondet C, Sonnendrucker E, Constantinesco A. Pinhole SPECT imaging : compact projection/backprojection operator for efficient algebraic reconstruction. IEEE Trans Med Im 2006, 25: 158-67 [2] Constantinesco A, Choquet P, Monassier L, Israel-Jost V, Mertz L. Assessment of left ventricular perfusion, volumes, and motion in mice using pinhole gated SPECT. J Nucl Med 2005; 46: 1005-11 [3] Arbogast-Ravier S, Xu F, Choquet P, Brunot B, Constantinesco A. Dedicated low field MRI: a promising low-cost technique. Med Biol Eng Comp 1995, 33: 735-9 [4] Loening AM, Gambhir SS, AMIDE: A free software tool for multimodality medical image analysis. Mol Imaging 2003, 2: 131-7

Speaker: Ms Elodie Breton (Service de Biophysique et Médecine Nucléaire, CHU Hautepierre, 1 av. Molière, 67098 Strasbourg, France)

14:00 Early Detection of Tumor Masses in Mice by In Vivo Hematoporphyrin mediated Fluorescence Imaging

In the last years we have tested hematoporphyrin (HP) dichlorohydrate as an optical contrast agent for the in vivo detection of solid surface tumors on small animals (mice) by Fluorescence Reflectance Imaging (FRI). HP dichlorohydrate was shown to accumulate in tumors, but the volume of the studied tumors was about 50-400 mm3, so that the tumor masses were visible with the naked eye and palpable. Instead, it is crucial in the tumor diagnostic the detection of tumors in their early stage: earlier diagnosis greater the patient survival probability, especially if the tumor is highly invasive. In this work, we apply HP-FRI to image in vivo the HP red fluorescence emission from mouse regions injected with tumor cells of high malignancy degree, in order to study the onset and the growth of the tumor and its early detectability. To this purpose, we improved our previous setup with a high sensitivity, high photometric resolution, monochrome, cooled digital CCD camera. The fluorescence excitation is supplied by a frequency doubled pulsed Nd:YAG laser (532 nm). The cooled digital CCD camera (Hamamatsu ORCA 285 G) records the HP fluorescence radiation (at 630 and 690 nm) filtered by a cut-on long-wavelength pass filter (cut-on wavelength = 600 nm) that rejects the backscattered radiation at 532 nm. About 106 ARO (anaplastic human thyroid carcinoma, highly malignant) cells were implanted subcutaneously in the back of several mice and their growing was monitored about daily through HP-FRI measurements for approximately 10 days starting from the fifth day after the cell injection and 6 h after HP injection. By exploiting the selective HP uptake by the tumor tissues and the high sensitivity of the CCD camera, we have imaged the fluorescence of tumors implanted only five days before the HP-FRI measurements, and, as a consequence, non visible with the naked eye and undetectable to the palpation. HP uptake appears different for healthy and tumor tissues so that the maximum optical contrast tumor-healthy tissue is obtained 24 h after HP administration. HP dichlorohydrate, through it is an aspecific fluorescent tumor marker, is suitable to detect subcutaneous tumor masses with a number of cells of the order of tens of millions (corresponding to a volume of few cube millimiters). Owing to the penetration depth in the tissues of emission red light of fluorophores exogenous (pheophorbide, probably), are detectable also some tumor-free internal organs. The features of this optical imaging system make it suitable for an integration with radionuclide imaging systems. Indeed, we combined it with Medipix2 hybrid pixel detectors (256x256 pixels of 55 micron pitch, 14x14 mm2 sensitive area) with silicon or CdTe pixel detector. A coded mask collimator with hole size of the order of 1-2 pixels would allow for a submillimetric spatial resolution. The resulting multimodal imaging system for small animals aims at reaching high spatial resolution (< 1 mm), short acquisition times, reliability, ease of use, compactness and relatively low cost. By taking planar views of the animal, the system can produce in vivo images of surface tumor masses grown in mice, using 99mTc MIBI as a radiotracer and hematoporphyrin (HP) dichlorohydrate as an exogenous optical fluorophore. Finally, the availability of radionuclide labelled porphyrins makes possible to administer at the same time both the gamma emitting radiotracer and the fluorescent marker, when aiming at a combined radionuclide / fluorescence imaging.

Speaker: Dr Maddalena Autiero (Università degli studi di Napoli "Federico II")

14:00 Experimental evaluation of the ClearPEM Detector Modules

ClearPEM is a project to develop a state of the art Positron Emitting Tomograph dedicated to early breast cancer detection. It is developed by a consortium of Portuguese research institutions within the framework of the Crystal Clear Collaboration. The scanner has a modular architecture, where each detector module features an array of 32 LYSO crystals, with double readout by two pixelated Avalanche Photo Diode arrays. The first detector modules have been successfully assembled and tested. In this talk we will present the results of these tests, in terms of Signal Yield, Photopeak Resolution, Asymmetry and Crosstalk.

Speaker: Ms Catarina Ortigão (LIP)

14:00 Implementation of a methodology for the analysis of small animal PET images

Introduction Small animal positron emission tomographs (PET) are becoming extremely popular allowing preclinical investigation of new drugs, new radiotracer etc. by performing several sequential scans on the same animal. In order to compare the results between scans (acquired at different time points) it is necessary to apply quantitative or semi-quantitative image analysis methods. True quantitative analysis can only be obtained by performing dynamic imaging combined with the measurement of the arterial input function (IF) in order to apply compartmental models. Such modelling technique is quite complex especially because of the intrinsic difficulties in measuring the IF. It is thus necessary to investigate alternative approaches. One possible choice is to measure the standard uptake value (SUV) defined as the ratio between the tissue radiotracer concentration and the injected activity per animal weight. Firstly in order to obtain accurate SUV values it is necessary to cross calibrate the scanner and to measure the amount of tracer in the animal tail. Estimation of the activity in the tail is quite important considering the intrinsic difficulties in performing good tail injections. Secondly it is necessary to develop a user friendly software tool that take into account all the calibrations including also the uptake time, reconstruction methods, animal weight etc. The objective of this work is to describe the calibration procedures and the software developed at our institution to obtain SUV images of small animals. Material and methods In order to estimate the tracer activity in the animal tail a set of calibrations were carried out using a 0.5 ml syringe (tail phantom) filled with a solution of 18- F ranging from 0.2 to 20 MBq. The coincidences per second (cps) were measured for all the three energy windows (EW) available on our system (GE eXplore Vista) and linear fits between the syringe activity and the total cps were performed. The parameters obtained from the fits (for each EW) were then used to estimate the tail activity from the total cps measured on the tail. The scanner cross calibration was performed by filling a mouse phantom (syringe of 26 mm diameter) with a 18-F concentration ranging from 0.3 to 1 MBq/ml. A set of images were acquired with three energy windows (100-700, 250-700 and 400-700 keV) and recostructed, after Fourier rebinning, using both FBP and OSEM. No corrections for attenuation or scatter were applied. Regions of interest (ROI) were drawn on the reconstructed images and linear fits between the images cps/ml obtained from the ROIs and the known mouse phantom concentrations were performed. All the results obtained using the calibration methods described above were implemented into a graphic user interface (GUI) code developed using IDL 6.2 (Interactive Data Language). The code allows the user to load an interfile image and to obtain as output an interfile “SUV image” that can be loaded into any image processing workstation. In order to evaluate the accuracy of the calibrations several measurements were carried out using three different phantoms. More precisely images of two cylindrical phantoms with different diameters respectively equal to 20 mm and 30 mm covering the entire axial field of view (46 mm) were acquired. The measurements were performed considering 18-F concentrations ranging from 0.3 to 0.9 MBq/ml. The measured radiotracer concentrations were then compared with the known true phantom concentrations in order to calculate the mean error. Secondly images of a custom made cylindrical phantom (with the same dimension as the mouse phantom) consisting in a hot sphere (10 mm diameter) floating into a uniform active background (sphere to background ratio equal to 3) were also acquired. The measured sphere 18-F concentrations were then compared with the true values (ranging from 0,4 to 2 MBq/ml). Results The correlation coefficients of the linear fits obtained using the tail phantom, mouse phantom and hot sphere phantom were always greater than 0.999, showing a good correlation between the measured cps/ml and the known true 18-F concentration. For the 100-700 keV EW the mean differences between the measured and true phantoms concentrations were respectively equal to 3.5 % and 2.2% for FBP and OSEM. The mean differences between the concentrations for the 250-700 keV EW were respectively equal to 5.5% and 4.7% for FBP and OSEM. For the 400-700 keV EW the mean concentrations differences were respectively equal to 5.4% and 4.5% for FBP and OSEM. Results obtained using the hot sphere phantom show that for FBP the mean differences between the measured and true hot sphere concentrations were respectively equal to: 12%, 7% and 2% for the 100-700 keV, 250-700 keV and 400-700 keV EW. For OSEM reconstruction the mean concentrations differences were equal to: 8%, 5% and 3% for the 100-700 keV, 250-700 keV and 400-700 keV EW respectively. Conclusions Results show that it is possible to obtain reasonably accurate calibrations and, thus, to obtain good estimate of radiotracer concentration in small animals. The mean errors were slightly higher when using the hot sphere phantom compared with the cylindrical phantoms, however in this case a much larger range of 18-F concentrations was considered. The developed code was also tested by several users showing good stability.

Speaker: Dr Antonello Spinelli (Policlinico S. Orsola-Malpighi, Bologna, Italy)

14:00 Initial experence in small animal imaging with clinical PET/CT scanner

Molecular Imaging is an emerging discipline that seeks to exploit our increased understanding of the molecular basis of disease through the design of novel imaging probes to specific molecular targets. Small animal PET offers the unique opportunity to in vivo image small animal models of human diseases noninvasively, repeatedly, and quantitatively in the same animal. Unfortunately, small animal PET scanner is not widespread and needs to additionally acquire anatomical image data (CT or MRI) to facilitate the precise target definition, which yields the need to seek for the alternative methods to deal with the problem. With the unique ability to acquire registered CT and PET images in the same scanner and provide clinical use and basic research, combined PET/CT scanner might be the potential option for small animal study. In addition, the CT-based attenuation correction can improve the overall visual quality and the quantitative accuracy of the radionuclide image data. Moreover, PET/CT images using a large-bore clinical scanner enables high- throughput studies to evaluate the performance of PET tracers in a timely and cost- effective manner by imaging multiple animals simultaneously. Dual-modality PET/CT imaging has had its great impact; however, the feasibility of PET/CT scanner which is developed for human imaging still needs to be concerned. Because of insignificant outcome and even producing much more noise, attenuation correction (AC) is not usually performed in microPET. In clinical PET/CT, it always adopts CT- based AC maps to improve image quality. However, the effect of AC technique on small animal imaging is still unknown. The main objective of this study is to investigate the effect of CT-based attenuation correction on performance of clinical PET/CT. We also proposed several CT imaging protocols in PET/CT system to investigate image quality of PET images and explore the performance difference between clinical PET/CT and microPET. The feasibility tests are taken on PET/CT scanner (Discovery LS; GE Medical Systems) and small animal PET (microPET R4) with costumed cylindrical and uniform phantom. Regarding the demand for single mouse and multiple mice image applications, different FOV settings were selected. Three different CT protocols, CT high resolution model (CT-H-RES), CT auto fusion model (CT-auto) and CT attenuation correction model (CT-AC) were employed. The scanning parameters for CT- H-RES were: field of view 15×15 cm2 for single rat scan 25×25 cm2 for multiple rat high throughput scan, matrix size 512×512, slice thickness/interval =0.625/0.625 mm, number of slice = 2, helical pitch = 1. For CT-auto, the parameters were: field of view 50×50cm2, matrix size 256×256, slice thickness/interval =5/4.25 mm, number of slice =4, helical pitch = 0.75. For CT-AC, the parameters were: field of view 15×15 cm2 for single rat scan; 25×25 cm2 for multiple rat high throughput scan; 50×50 cm2 for CT-auto model, matrix size 512×512, slice thickness/interval =5/4.25 mm, number of slice = 4, helical pitch = 0.75, and coverage = 150 mm covering the whole phantom or rat. We use 80 kV and 300 mA for all scanning protocols. A cylindrical and uniform phantom was used to evaluate the performance of PET/CT including resolution, uniformity, and signal to noise ratio. The cylindrical hot spot phantom contains 4 by 4 cylindrical matrix in different bore size: 3, 2.5, 2, 1.75, 1.5, 1 mm to validate clinical PET/CT resolution level. Another cylindrical hot spot phantom containing various bore size from 2.6 to 1.2mm was used to investigate the resolution characteristics of microPET. To evaluate uniformity and signal to noise ratio of clinical PET/CT and microPET scanner, we constructed a homogeneous phantom at small animal size consisting of a hollow acrylic cylinder, 5.8 cm in inner diameter and 2.2 cm in height. In this study, we also inspect the intrinsic spatial resolution based on NEMA NU-2001 standard (National Electrical Manufacturers Association). The full width at half maximum (FWHM) and full width at tenth maximum (FWTM) are used to characterize the spatial resolution. The PET images in resolution, uniformity, and signal to noise ratio under our protocol are evaluated. In single animal scan with CT-based attenuation correction: the spatial resolution = 2.5mm, signal to noise ratio = 27, Uniformity = ±8.5%; compared to without attenuation correction: the spatial resolution = 2.5mm, signal to noise ratio = 13, Uniformity = ±15.6%. In multiple animals scan with CT-based attenuation correction: the spatial resolution = 3mm, signal to noise ratio = 29, Uniformity = ±7.4%; compared to without attenuation correction: the spatial resolution = 3mm, signal to noise ratio = 15, Uniformity = ±14.3%. In the without CT-based AC protocol, we observed that resolution and signal to noise level did not reduce significantly, but image uniformity did reduce obviously. The resolution, uniformity and signal to noise ratio of microPET image is about 1.8mm, 8.5 %, and 27, respectively. Following NEMA NU-2001 in spatial resolution, the resolution is characterized by the FWHM and the FWTM levels of the reconstructed PSF in the 3 orthogonal directions (e.g., radial, tangential, and axial direction). This allows a best-case evaluation of scanners, taking into account the variation in resolution with radial distance which was chosen by different purpose with FOV. Here we summarized the transverse and axial resolution at various positions. Spatial resolution represented as FWHM at center of scanner: Axial resolution is 14.64mm, Transverse resolution is 5.49mm; and represented as FWTM: Axial resolution is 21.41mm, Transverse resolution is 8.95mm. At off center of scanner about 7.5cm which is to indicate side resolution at 15×15 FOV which represented as FWHM: Axial resolution is 15.14mm, Transverse-radial resolution is 5.91mm, Transverse- tangential resolution is 5.79mm; and represented as FWTM: Axial resolution is 21.93mm, Transverse-radial resolution is 9.31mm, Transverse-tangential resolution is 9.55mm. At off center of scanner about 12.5cm which is to indicate side resolution at 25×25 FOV which represented as FWHM: Axial resolution is 17.57mm, Transverse-radial resolution is 6.65mm, Transverse-tangential resolution is 5.80mm; and represented as FWTM: Axial resolution is 20.29mm, Transverse-radial resolution is 6.65mm, Transverse-tangential resolution is 9.27mm. In this study, we have found that the overall image quality from the microPET scanner is superior to those from the PET component of PET/CT scanner due to smaller crystals and smaller detector ring diameters. Note that the difference is not quite significant. From the scanning protocol aspect, we observed that the image quality under the multi-mice scanning protocol was as well as those under single-mouse imaging protocol suggesting that high-throughput studies may be feasible with the reliability we demand. However, if high resolution images are vital, the multi-mice scanning protocol was not highly recommended to use because of the relatively lower resolution resulting from the off-center effect and relatively large FOV. Considering the effect of CT-based attenuation correction, we have found that corrections for photon attenuation are important to improve the image uniformity characteristics either in single-mouse or multi-mice protocol setting. In summary, although clinical PET/CT scanner can not compete with microPET scanner especially in spatial resolution due to the intrinsic geometry designs, the combined CT can bring about several possibilities in research application such as high-throughput studies. In addition, the precise fusion of molecular PET images with high-quality anatomical CT images facilitates anatomic localization of the PET findings, overcoming alignment problems due to internal organ movement, variations in scanner bed profile, and positing of the patient for the scan, which is often encountered with techniques that register images obtained from 2 separate systems. Furthermore, the CT-based attenuation correction can improve the recognition of radionuclide imaging and potentially the quantitative accuracy of the reconstructed image. Finally, considering all the advantages, clinical PET/CT imaging might be potential technique for small animal study.

Speaker: Ms Chia-Lin Chen (National Yang-Ming University)

14:00 Method for subject recognition (identification)

New method for subject recognition is proposed. We need to know sets of several parameters of assumed subjects for realization of this method. On the base of knowledge about value of these parameters and them distribution characteristics we can say whether investigated subject is assumed subject or no. Parameters of proposed method can have discontinuous or continuous value. In case of discontinuous parameters we use logic approach for problem-solving of identification of subjects. Than we have two types of results: (1) investigated subject is assumed subject; or (2) investigated subject is not assumed subject. In case of continuous parameters we use mathematical statistics approach for problem-solving of identification of subjects. Than we have continuous set of results corresponding probability that whether investigated subject is assumed subject or no. Reliability of subject recognition depends on number of using parameters. Proposed method can be used for different applications. For example, it can be used for identification of kind of wood by contain of trace contaminant into ashes; it can be used for medical diagnostics and identification of type of diseases by several factors; it can be used for pattern recognition and so on. So proposed method can found of wide application.

Speaker: Prof. Alexander Tcherepanov (senior lecturer)

14:00 Method to localize a fluorescent inclusion in a turbid medium using time-resolved systems

Optical fluorescent tomography can provide information about the spatial distribution, the lifetime and the concentration of specific fluorescently-labeled cell markers, inserted in highly scattering biological media. 3D fluorescent optical tomography systems are currently developed either with continuous wave, or in the frequency domain or in the time domain. The latter is the one which contains the richest information for a signal obtained from a single source-detector. A method to solve of the inverse problem is to extract some parameters from the whole temporal curve and to reconstruct the problem from these parameters. The aim of this work is to present an analytical approach using the mean time of the temporal signal to determine the position of a fluorescent inclusion in a turbid medium. We tested our method on optical phantoms on two experimental systems, a Time Correlated Single Photon Counting and a gated CCD camera. Light propagation is assumed to satisfy time-dependent diffusion equation. Analytical solutions describing the fluorophore response are expressed under several approximations. The general solution for fluorescence is the double convolution of the solution at the excitation wavelength for the propagation from the source to the inclusion by the fluorescence decay lifetime and then by the solution at the emission wavelength for the propagation from the inclusion to the detector. Analytical expression for the moment is obtained from derivations of expressions of the photon density in the frequency domain. Because they connect localization of the inclusion with the source-detector positions and optical properties of the medium, we can recover the localization of the inclusion from a set of experimental acquisitions.

Speaker: Dr Anabela da Silva (Cea-Leti Recherche Technologique)

14:00 Miniaturized Fibered Confocal Microscopy with the Cell~vizio now enables in vivo in situ imaging

The development of the Red Cell~vizio, using a red 635 nm laser, makes it possible to obtain microscopic in vivo images even with the large range of red fluorescent dyes that are widely used in the field of molecular imaging. The Cell~vizio™, based on Fibered Confocal Fluorescence Microscopy, enables in vivo and in situ microscopy down to a resolution of 2.5 µm. Until now, the Cell~vizio™ was only available with a blue 488 nm laser catering to the range of fluorophores and transgenic animals that have an emission bandwidth of 500 to 650 nm. The introduction of the Red Cell~vizio now offers the possibility to cover a large number of fluorophores like Cy5.5 or animal models that emit fluorescent signal between 650 and 750 nm. Such wavelengths are being widely used in whole body optical imaging techniques and hence the Red Cell~vizio allows the use of the same models to obtain microscopic in vivo real-time images. The Cell~vizio is a stand-alone imaging system made of a Laser Scanning Unit, a range of fibered objectives made of tens of thousands of optical fibers (the “ProFlex”) and the ImageCell image processing and analysis software. High definition images with 2.5 µm lateral resolution, 15 to 20 µm axial resolution, and a field of-view up to 600 µm x 500 µm can be obtained at a rate of 12 frames per second on living anesthetized animals. These images can be obtained at depths up to 100 µm. The fibered probes have diameters ranging from 300 µm to 1.8 mm and thus offer unique access capabilities. The Cell~vizio has been applied to scientific research across the world in the areas of peripheral nerve studies, deep brain imaging, microcirculation, angiogenesis, immunology, oncology and ophthalmology. Convincing results and images were obtained in those fields.

Speaker: Ms Anne Osdoit (MKT)

14:00 Molecular imaging through 1H MRS and MRSI in everyday routine: improvements in various clinical applications and parameter optimization of spectroscopic imaging sequences.

In the era of molecular imaging, in-vivo 1H Magnetic resonance spectroscopy (MRS) and spectroscopic imaging (MRSI) are impacting dramatically upon virtually all areas of clinical medicine. MRS and MRSI should be able to identify key biochemical changes, much before the tumour becomes detectable by other functional imaging methods that mainly rely upon single markers that are not entirely sensitive or specific for malignant activity. Combined with other imaging techniques in a rapidly advancing modality like MRI offer the ability to estimate the presence of metabolites yields much information regarding tissue. Molecular imaging through magnetic resonance could be potentially suited for screening and repeated monitoring since it entails no exposure to ionizing radiation. Incorporation of these tools in clinical practice is, however, limited due to the considerable amount of user intervention. In this work various acquisition parameters and their effects in spectrum quality are investigated. In order to assess the quality of various spectroscopic techniques (2D and multi-slice MRSI, multiple echo SI), a series of experiments were conducted using a standard solution. The application of water and fat suppression techniques and their compatibility with other parameters were also investigated. The stability of the equipment, the appearance of errors and artifacts and the reproducibility of the results were also examined to obtain useful conclusions for the interaction of acquisition parameters. All the data were processed with specialized computer software (jMRUI 2.2) to analyze various aspects of the measurements and quantify various parameters such as signal to noise ratio (SNR), full width at half maximum (FWHM), peak height and j-modulation. The experience acquired from the conducted experiments was successfully applied in acquisition parameter optimization and improvement of clinical applications (2D MRSI of prostate, brain and muscle MRS) by significantly improving the spectrum quality, SNR (up to 75%), spatial resolution in 2D MRSI, water and fat suppression and in some cases reducing exam times (up to 60%).

Speaker: Mr Anastasios Karatopis (Department of medical instrumentation,Technological educational institution/2nd Department of Radiology, Medical School, University of Athens)

14:00 Molecular Switch of Cre/loxP for Radiation Modulated Gene Therapy on Hepatoma

Alpha-fetal protein (AFP) is a specific tumor marker for hepatocellular carcinoma (HCC). The gene expression system under AFP promoter/enhancer control would be specific for AFP producing cells, but its low expression level is a problem to be overcome. For the purpose of AFP promoter enhancement for the use of radiation modulated gene therapy, we combined hepatitis B virus (HBV) enhancer II with AFP promoter which shows the selectivity to target cells to control the Cre/loxP system. Different gene constructs, such as pE4luc, pE4Tk, EIIAPA-Cre, E4CMV-STOP-Tk and chimeric promoters combined with HBV enhancer were constructed. HepG2, HeLa and NIH- 3T3 cell lines were then transfected by using jet PEITM reagent. Luciferase assay and MTT technique were used to analyze the luciferase gene expression and Herpes simplex virus thymidine kinase (HSV1-Tk) gene function, respectively. Cells irradiation experiments with a dose range of 1~10 Gy were performed by a 60Co gamma irradiator. HSV1-Tk gene expression of stable clone implanted on the NOD/SCID mice was assessed by in vivo 131I-FIAU imaging. The E4 enhancer showed better response to radiation after 60 hours irradiation especially at a dose range of 5~7 Gy in pE4luc HepG2 stable clone. The EIIAPA promoter provides high specificity to hepatoma but not to non-hepatoma cell lines. It also activates the Cre downstream and removes the stop cassette only in hepatoma cells. After removal of the stop cassette, the E4 response to radiation could encode more Tk protein and kill more tumor cells. In summary, the chimeric EIIAPA promoter can stringently control the expression of Cre recombinase only in HCC, and the removal of stop cassette let the HSV1-Tk gene expresss and subsequently kill the tumor cells. The radiation effect of the EIIAPA-Cre and E4CMV-STOP-Tk system shows promising results in terms of cell survival of HCC.

Speaker: Ms Ya-ju Hsieh (Institute of Radiological Sciences, National Yang-Ming University, Taiwan)

14:00 Multi-Energy Densitometry of Bone Tissues Using ZnSe-based Scintielectronic Detectors

Modern digital radiography and X-ray computer tomography (CT) are among the most efficient methods used for diagnostics of cancer and other diseases requiring surgery. Transition from traditional to dual-energy CT allows separate diagnostics of bone and muscle tissues, i.e., discern between biological materials that substantially differ in their density and atomic number Zeff (by 2-3 times). For early detection of tumors, atherosclerosis and osteoporosis, it is essential to detect, by a non-invasive method, deviations of these physical parameters by 15-20% from their normal values. In the particular case of hepatitis, it is required to determine with such accuracy the content of calcium in the bone tissue. A vital factor for solution of this problem is selective energy sensitivity of high- and low-energy detectors. In traditional dual-energy radiography [1], the low-energy detector is made of gadolinium oxysulfide, and the high-energy detector - CsI(Tl). For diagnostics of early stages of osteoporosis, we propose to use an original combined detector [Patent of Ukraine No. 44547 of 15.02.02], where we use, for detection of low energies, scintillator ZnSe doped with an isovalent admixture (Cd, Te, O). This exclusive material is produced solely by ISMA of STC "Institute for Single Crystals", Kharkov, Ukraine. Light output of ZnSe is 20-30% higher as compared with CsI(Tl), and the afterglow level is 10-2 % after 3 ms. Another peculiar feature of ZnSe is its relatively low Zeff = 33. Thus, it is nearly not sensitive to high energies (above 100 keV) when used in the form of thin detectors (0.3-0.5 mm), with its maximum detection efficiency corresponding to the low energy range (30-40 keV). As a result, energy selectivity of the dual-energy system to high energies is ensured automatically. Another advantage is that placing this crystal before the high-energy detector ensures efficient filtration from the low-energy part of X-ray spectrum. This lowers the requirements to the filter (usually made of copper with Zeff = 31) or makes its use unnecessary. The use of zinc selenide crystals in state-of-the art inspection equipment [2] has already led to excellent results in detection of explosives in X-ray inspection of loads and luggage. Our comparative experimental studies of animal bone tissue samples in-vitro for determination of calcium composition in combination with tomography have confirmed high sensitivity and detecting ability of the method. The accuracy of determination of the effective atomic number reaches 80-90%. The sensitivity to relative variations of bone tissue density is at the level of 1-2%. Moreover, quality of radiographic images is improved, which were obtained using a dual-energy tomographic scanner in sandwich design of detectors 0.5 mm ZnSe / 4 mm CsI(Tl) / Hamamatsu photodiode. The dynamic range is 10^6, the contrast to noise ratio is of the order of 2-4%. Further studies will be aimed at separate visualization of trabecular and cortical bone tissues and determination of changes in their chemical composition related to the effects of different stages of osteoporosis. In addition, further development of multi-energy tomography on the basis of modern bright and fast scintillation materials, including ZnSe compounds, is expected to enable detection of microcalcification of soft tissues and organs, as well as diagnostics of tumor formations by means of mammography, etc. [1] R.M. Harrison, NIM 310, 24 (1991). [2] Smiths Heimann Prospects, 2006, http://www.heimannsystems.com

Speaker: Prof. Vladimir Ryzhikov (Institute for Scintillation Materials of STC "Institute for Single Crystals")

14:00 PIXSCAN: Pixel Detector CT-Scanner for Small Animal Imaging

The PIXSCAN is a small animal CT-scanner based on hybrid pixel detectors. These detectors provide very large dynamic range of photons counting at very low detector noise. They also provide high counting rates with fast image readout. Detection efficiency can be optimized by selecting the sensor medium according to the working energy range. Indeed, the use of CdTe allows a detection efficiency of 100% up to 50 keV. Altogether these characteristics are expected to improve the contrast of the CT-scanner, especially for soft tissues, and to reduce both the scan duration and the absorbed dose. A proof of principle has been performed by assembling into a PIXSCAN prototype the photon counting pixel detector initially built for detection of X-ray synchrotron radiations. Despite the relatively large pixel size of this detector (330 x 330 µm2), we can present nice tomographic reconstruction of mice at good contrast and spatial resolution. A new photon counting chip (XPAD3) is designed in sub-micronique technology to achieve 130 x 130 µm2 pixels (see the poster "XPAD3: A new photon counting chip for X-ray CT-Scanner " at this conference). This improved circuit has been equipped with an energy selection circuit to act as a band-pass emission filter. Such energy selection should improve the image quality by cutting the low and high energy queues after the pre-filtered X-ray source. For example in the presence of iodinated contrast agents, one can select an energy window lower and higher than the iodine photoelectric absorption jump and, by image subtraction, have the iodine only. Furthermore, the PIXSCAN XPAD3 hybrid pixel detectors will be combined with the Lausanne ClearPET scanner demonstrator (see poster "Measured and Simulated Specifications of the Lausanne ClearPET Scanner Demonstrator" at this conference). CT image reconstruction in this non-conventional geometry is under study for this purpose.

Speakers: Prof. Franck Debardieux (Institut de Biologie du Developpement de Marseille, France), Dr Pierre Delpierre (Centre de Pysique des Particules de Marseille)

Paper

• PIXSCAN_Figures_poster95_delpierre_debardieux.doc
• PIXSCAN_Figures_poster95_delpierre_debardieux.pdf
• PIXSCAN_paper_poster95_delpierre_debardieux.doc
• PIXSCAN_paper_poster95_delpierre_debardieux.pdf

14:00 Polysialylation increases lateral mobility of Neural Cell Adhesion molecule (NCAM) in the cell membrane.

Polysialic acid (PSA) is a long, linear a2-8 linked carbohydrate of N-acetylneuraminic acid residues, added post-translationally to the Neural Cell Adhesion Molecule (NCAM), the prototype of a large family of cell surface Immunoglobulin-like Super-Family (IgSFs) expressed in the nervous system. Polysialylation of NCAM is thought to facilitate cell migration and plasticity by inhibiting cell adhesion and interactions with extracellular matrix as a result of the large excluded volume of the PSA polymer. Here, using real time live cell imaging techniques (i.e. Time lapse, Fluorescence Correlation Spectroscopy and Fluorescence Recovery After Photobleaching) we uncovered an action of PSA in regulating specifically the tune of NCAM lateral diffusion. PSA increases the rate of lateral diffusion of NCAM mobile fraction in the cell membrane without affecting the proportion of the mobile fraction or the nature of NCAM’s diffusion. These findings reveal a new molecular mechanism by which polysialylation regulates NCAM dynamics.

Speaker: Dr Valéry Matarazzo (IBDML, CNRS-UMR6216, Université de la mediterannée)

14:00 Preliminary evaluation of the tomographic performance of the MediSPECT small animal imaging system

We report on the tests of a prototype (MediSPECT) for SPECT imaging on small animals with a small field of view (FOV) and high spatial resolution. MediSPECT is a SPECT imaging system based on a 1-mm thick CdTe pixel detector bump-bonded to the Medipix2 CMOS readout circuit operating in single-photon counting developed by the European Medipix2 collaboration. The CdTe detector has 256x256 square pixels, 55 micron by side, for a sensitive area 14 mm x 14 mm. In its present version, this system has a single detector head mounted on a rotating gantry. A detailed description of MediSPECT is reported in a companion work presented at this conference. For preliminary testing and calibration of the acquisition equipment and image reconstruction algorithms, 90 projections of a gamma-ray point source (403 kBq Cd-109) through a single pinhole (diameter 0.75 mm; radius of rotation about 2.5 cm; focal length about 4.5 cm) were acquired in a step-and-shoot mode. A sensitivity of 4 cps/MBq (at 22 keV) was measured. Capillaries 800 micron in diameter were arranged in a Y shape to form a more complex phantom (45 projections, about 1 mCi I-125). Images were reconstructed with a custom algorithm implementing OSEM. Given the high resolution that the system seeks to ultimately achieve, the code incorporates center of rotation correction. Results show successful image reconstruction of these simple phantoms. In the near future, we expect to be able to image more complex phantoms as a preliminary step to further validate the methods so far developed for the long-term goal of high- resolution (<~300 micron) in vivo imaging over a small field of view with single as well as multiple pinhole optics.

Speaker: Dr Maria Cristina Montesi (Dipartimento di Scienze Fisiche, Università di Napoli Federico II, and INFN, Sezione di Napoli, I-80126 Napoli, Italy)

14:00 Preliminary Report on LabPET, a High-Performance APD-Based Digital PET Scanner for Small Animal Imaging

LabPET is a second-generation APD-based PET scanner featuring quasi-individual crystal readout and massively parallel digital processing with advanced signal analysis heuristics for high-performance in vivo molecular imaging of small animals. The system implements several unique features to enhance throughput and facilitate complex imaging protocols in live animals. The basic LabPET system consists of 3072 2x2x10 mm³ scintillators assembled into phoswich LYSO/LGSO pairs, each read out by an avalanche photodiode. The detectors are arranged into 16 rings of 192 crystals each, having 16.2 cm in diameter by 37.5 mm axially. The 16 rings allow for acquisition of 31 effective image slices extending up to 110 mm in diameter. The dual-crystal readout scheme avoids resolution degradation due to light- or charge-sharing and enables very high singles count rates with low dead time in the detector front-end. Advanced parallel digital signal processing and analysis algorithms are implemented in high-performance programmable devices for crystal identification, energy discrimination and time stamping of events from individual pixels. Virtually error-free crystal identification is achieved using an auto-regressive least-mean square moving average method. Coincidence events are sorted in real time and recorded in list mode, together with scanner status information and animal physiological data. The measured intrinsic spatial resolution of coincidence response functions is 1.2 mm FWHM and 2.3 mm FWTM both axially and radially, with a near triangular shape confirming the excellent resolving capability of APD detector arrays. Using a simple linear interpolation algorithm, preliminary time resolutions of 4.5, 7.5 and 9.6 ns FWHM were obtained for LYSO-LYSO, LYSO-LGSO and LGSO-LGSO coincidences, respectively. These resolutions will be improved by implementing digital constant fraction or artificial neural networks algorithms. Initial images with a resolution phantom show that 1.2 mm hot rods can be clearly resolved. The scanner is fully integrated with ancillary devices (injectors, life-sign monitors, blood counter, etc.), all controlled through a centralized user-interface, allowing a single operator to perform elaborate imaging protocols with maximum flexibility and efficiency. The system was designed to provide a fully integrated solution for performing the most demanding molecular imaging investigations with increased throughput in a busy research environment.

Speaker: Prof. Roger LECOMTE (Université de Sherbrooke, Nuclear Medicine & Radiobiology)

14:00 RECENT TECHNOLOGIC DEVELOPMENTS ON HIGH RESOLUTION BETA IMAGING SYSTEMS FOR MOLECULAR IMAGING WITH MULTI-ISOTOPES LABELING APPLICATIONS

Two novel betaimaging systems, particularly interesting in the field of radiopharmacology and molecular biology researches, were developed these last years: (1) a betaimager was derived from researches conducted by Pr Charpak at CERN. This parallel plate avalanche chamber is a direct detection system of  radioactivity, which is particularly adapted for qualitative and quantitative autoradiography. With this detector, autoradiographic techniques can be performed using classical pure emitters like 3H, 14C, 35S, …but also radionuclides, such as 99mTc, which are  emitters associated with low energy electrons while decaying. In fact the detector has a very low sensitivity to low-range  rays and only detects the electrons. The performance in the field of sensitivity (smallest activity detected: 0,007 cpm/mm2 for 3H and 0,01 for 14C), linearity (over a dynamic range of 104) and spatial resolution (50m for 3H or 99mTc ) gives a real interest to this system as a new imaging device. Its principle of detection is based on the analysis of light emitted during the interaction with an intensified CCD camera. (2) a micro imager is based on contact imaging through a solid scintillator sheet. Light emitted is amplified through an image intensifier tube and is analysed with a CCD camera. The full field of view is smaller than the first one (24mm x 32mm versus 20cm x 25cm) but a better spatial resolution is obtained (typically 15 m for 3H and 99mTc). The specifications of this detector are similar to the first ones with a slightly reduced sensitivity. Using these detectors, quantification is much easier and more precise than that of the radiological film because of direct counting of radioactivity. With new recent developments on these devices, it is possible to detect, in the same sample, two isotopes of different energies ( 3H and 14C for exemple ) or of different decaying periods ( 99mTc and 201Tl for exemple). Results obtained on biological experimentations for the development of different radiopharmaceuticals and interest of this new methodology in various applications, among which detection of positive electrons ( +: like those emitted by 18F, 15O or 11C ) will be detailed in this presentation.

Speaker: Dr Nicole BARTHE (INSERM U577, Université Victor Segalen Bordeaux 2)

14:00 Results obtained with the ClearPET(TM) Rodent small animal PET scanner

We will report result obtained with a ClearPET(TM) Rodent small animal PET scanner at the Vrije Universiteit Brussel (VUB). The ClearPET(TM) scanner was build to be modular. At its basis it consists of cassettes that can contain up to four 64 channel PMTs (Hamamatsu R7600-M64) in line. The scintillator used is a combination of LYSO and LuYAP (each 8mm long), that can be distinguished through their different decay times. This effectively gives DOI information. The signals are digitised using free running ADCs operating at 40MHz. The cassettes are mounted on a rotating gantry, allowing the use of a partial ring geometry. The scanner under test comprises of 8 cassettes each equipped with 2 PMTs. Due to an axial shift introduced into every 2nd cassette, the effective axial field of view amounts to 50mm. The cassettes are placed such that each cassette has another cassette directly opposing it (180° further at the gantry). Measurements of the spatial resolutions and sensitivity will be presented and compared with simulations obtained with Gate. Furthermore, a position calibration method will be outlined and the influence on the spatial resolution will be quantified. Due to the peculiar geometry, a dedicated image reconstruction procedure based on direct list mode reconstrution was developped and will be presented. Images with a Derenzo phantom and first (18)FDG images with rats will also be shown, illustrating the potential of the scanner.

Speaker: Dr Olivier Devroede (Vrije Universiteit Brussel)

14:00 Scintillator studies for the HPD-PET

In the framework of the feasibility studies of the 3D Axial HPD-PET, consisting of axially oriented arrays of long (Lc from 100 to 150 mm) scintillation crystals of small cross section (3.2 x 3.2 mm2) read-out on both sides by a Hybrid Photon Detector (HPD), various investigations are being performed on the features of the crystals to be employed. In fact, the axial spatial, energy, and time resolutions of such a device are influenced by No and Leff, the key parameters of the HPD-PET concept. No is the number of photoelectrons produced by a 511 KeV gamma-ray in a crystal of short length (Lc→ 0). Its value depends both on the physical and optical properties of the chosen scintillator (including its surface finishing and coating/wrapping) and on the characteristics of the photodetector. Leff is the effective light attenuation length of optical photons on their way inside the crystal from the scintillation point to the photodetectors. While an increase of No improves all the resolutions, an increase of Leff, although improving energy and time resolutions, worsens the axial one. We will present the experimental study carried out to optimize the effective light attenuation length, Leff, for a set of polished YAP:Ce scintillators of dimensions 3.2×3.2×100 mm3 produced by Preciosa Crytur Co at Turnov, Czech Republic, and of polished LYSO:Ce of equal size, produced by Photonic Materials, Bellshill, Scotland. We will discuss the resolutions that have been measured for the various Leff values obtained in polished and wrapped crystals at 511 keV gamma-rays by means of standard photomultipliers with borosilicate windows and bialkali photocathodes. Even if the choice of the scintillator-wrapping to be used in the final project has not yet been made, these values, extrapolated for the HPD photodetectors, that will be equipped with a thin sapphire entrance window, give hints on the capabilities of the novel 3D Axial HPD-PET concept. The possible perspectives will also be discussed.

Speaker: Raffaele DE LEO (PHYSICS Department Bari University)

14:00 Sensitivity of limited angle TOF PET systems

Due to the availability of scintillators (LSO, LaBr3) with good timing resolution, sufficient stopping power and good energy resolution, different 3D TOF PET scanners are now under development. Recently we have shown that improved timing resolution allows to reduce the number of angles needed to reconstruct TOF-PET data. Histoprojections from TOF data can be seen as an asymmetrically blurred image with noise. Therefore image reconstruction from histoprojections can be seen as a problem of image restoration (earlier discussed in a paper by Grangeat) and this reconstruction method was implemented to reconstruct the data. Using these reconstruction techniques a full ring configuration is not needed to allow reconstruction. As it has been shown before for non-TOF PET systems a full ring configuration is not optimal from a sensitivity perspective for a given number of detectors. This has led to development of half ring systems like the ECAT ART. The disadvantage of these PET systems was the need for rotation to obtain complete data. This is however not needed for limited angle TOF PET systems. Methods and results We quantified the changes in sensitivity and scatter fraction between different configurations by Gate Monte Carlo simulations. A full ring system with either 60 rings or 120 rings of LaBr3 crystals (4x4x30 mm) was modeled. A limited angle system with the same number of detectors as the 60 ring system was constructed by only using half of the ring. Different activity distributions where simulated to determine the change in sensitivity, Most objects for PET imaging will at least cover the whole axial FOV of the camera. An 5cm radially off center line source was used as a representative source for the volume sensitvity. The sensitivity increases for a full 120 ring system compared to the 60 ring system with 298 %. The semi ring (two opposed curved detectors) with the same amount of detectors as a full ring system has 99% more sensitivity. For a line source in a 70 cm long phantim the scatter fraction increases with 3 % when going from the complete 60 ring system to a limited angle 120 ring system. For small objects (less than about 25 cm axial extent of the 60 ring system) the situation is different. Here we used a central point source for estimating the sensitivity. When the number of rings of a complete ring system is doubled, the point source sensitivity increases by a factor of 99.5 %. So there is not an increase in sensitivity for a limited angle system with the same number of detectors. For small objects it is however possible to reduce the radius of the two parts of the limited angle system. We have quantified this increase in sensitivity. Conventional ring systems have a radius of about 45 cm. When we reduce the radius of a limited angle system to 40 cm we gain 34 % sensitivity, at 35 cm it is 86 % and at 30 cm the increase it 103%. Conclusions: A limited angle system with the same number of detectors has a doubled volume sensitivity and the same point source sensitivity. By modifying the radius of the scanner for small objects the sensitivity can be increased. This makes it possible to use a limited angle TOF PET system for brain or breast imaging without the disadvantage of conventional whole body PET systems.

Speaker: Dr stefaan Vandenberghe (Ugent)

14:00 Shrunken detection volumes in fluorescence 4Pi-microscopy

Keywords: Fluorescence Confocal Microscopy, 4Pi-microscopy, Molecular Detection Efficiency Function, Lateral Resolution, Axial Resolution. Improving the spatial resolution in optical microscopes is a challenging task for many applications. In fluorescence confocal microscopy the resolution is given by the Detection Efficiency Function (DEF) resulting from the product of the Excitation Efficiency Function (EEF) by the Collection Efficiency Function (CEF). The focalisation of the pump laser beam by a high numerical aperture objective lens defines the excitation efficiency volume whereas the image of the pinhole represents the collection efficiency volume. The axial dimension of the detection efficiency volume of fluorescence microscopes can be considerably improved by superposing two coherent illumination beams [1] and by adding coherently the two fluorescence wave- fronts emitted on the both sides of the luminescent sample [2]. This solution implemented in 4Pi-microscopes [3] must be coupled with a two-photon excitation mode [4] to be really efficient along the optical axis. Unfortunately, this method increases the lateral dimensions of Detection Efficiency Volume because of the infrared wavelength used for the illumination. Within this context, we propose an arrangement of the 4Pi-microscope [5] using the “spatial incoherence” of fluorescent samples to shrink the collection efficiency volumes by a factor of two [6]. In this case, the lateral extent of the Detection Efficiency Function is smaller with a two- photon [7] excitation mode than the one obtained in the classic case, with a one- photon excitation mode. Moreover the amplitudes of the side lobes are strongly reduced. References: 1. C. J. R. Sheppard and Yunrui Gong, “Improvement in Axial Resolution by Interference Confocal Microscopy,” Optik 87, 129-132 (1991). 2. N. Sandeau, H. Giovannini, P.-F. Lenne and H. Rigneault, "Observation of the interferences between the emitted beams in a 4Pi microscope by Partial Coherence Interferometry" Applied Physics Letters 87 (18103), 2005. 3. S. Hell and E. H. K. Stelzer, “Properties of a 4pi Confocal Fluorescence Microscope,” J. Opt. Soc. Am. A 9, 2159-2166 (1992). 4. S. Hell and E. H. K. Stelzer, "Fundamental improvement of resolution with 4Pi- confocal fluorescence microscope using two-photon excitation," Opt. Comm. 93, 277- 282 (1992). 5. S. Hell, European Patent Application EP0491289 (filed 18. dec. 1990) published 1992. 6. N. Sandeau and H. Giovannini, “Increasing the lateral resolution of 4Pi fluorescence microscopes” accepted for publication in J. Opt. Soc. Am. A (Nov. 2005) 7. N. Sandeau and H. Giovannini, “Arrangement of a 4Pi microscope for reducing the confocal detection volume with two-photon excitation” submitted to Optics Communications (Nov. 2005)

Speaker: Dr Nicolas Sandeau (Institut Fresnel)

Paper badge.pdf test

14:00 Small Animal and online hadrontherapy PET detector characterisation

Our group at the “Institut de Physique Nucléaire de Lyon” (IPNL) is working for physics and detectors for medical imaging. We are presently developing, for a demonstration purpose and within the Crystal Clear international collaboration, a small animal PET scanner that should have enhanced slow control and data acquisition features. We also investigate a feasibility study of an online PET dedicated for inline and in-situ dose deposition control in hadrontherapy. Our basic block detector is a compound of a single block scintillating crystal whose response is sampled by a multi-anode photomultiplier with its proper embedded acquisition readout chain. After a presentation of the characteristics of our basic experimental setup, we discuss our calibration algorithm that allows energy, impact point and depth of penetration measurement with good accuracy within the crystal block. Moreover, the algorithm is iterative, easily implemented and dependent from the number of sampling photomultiplier anodes a sub-millimeter resolution in all of the three coordinates is achievable.

Speaker: Mrs Marina TAVERNE (IPNL groupe imagerie médicale)

14:00 Temporal Monitoring and Quantitative Assessment of Tumor Burden Growth Using In Vivo Bioluminescence Imaging

In vivo bioluminescence imaging (BLI) is a sensitive imaging modality that is rapid and accessible, and may comprise an ideal tool for evaluating tumor growth. In this study, the kinetic of tumor growth has been assessed in C26 colon carcinoma bearing Balb/C mouse model. The ability of BLI to noninvasively quantitate the growth of subcutaneous tumors transplanted with C26 cells genetically engineered to stably express firefly luciferase and herpes simplex virus type-1 thymidine kinase (C26/tk- luc). A good correlation (R2 = 0.998) of photon emission to the cell number was found in vitro. Tumor burden and tumor volume were monitored in vivo over time by quantitation of photon emission using Xenogen IVIS 50 and standard external caliper measurement, respectively. At various time intervals, tumor-bearing mice were imaged to determine the correlation of in vivo BLI to tumor volume. However, a poor correlation of BLI to tumor volume was observed when tumor volume reached about 1000 mm3 (R2 = 0.907). Gamma scintigraphy combined with [131I]FIAU was another imaging modality used for verifying the previous results. In conclusion, this study showed that bioluminescence imaging is a powerful and quantitative tool for the direct assay to monitor tumor growth in vivo. The dual reporter genes transfected tumor-bearing animal model can be applied in the evaluation of the efficacy of new developed anti-cancer drugs.

Speakers: Ms Chia-Chi Chen (Cancer Research Division, National Health Research Institute, Zhunan, Taiwan), Ms Ya Fang Chang (Institute of Radiological Sciences, National Yang-Ming University)

14:00 The SPECT Program at IASA

The experimental program of the SPECT group at the Institute of Accelerating Systems & Applications (IASA) will be presented. It is mainly focused on the development of a new small field of view gamma-Camera based on a Position Sensitive Photomultiplier Tube with pixelized CsI(Tl) and NaI(Tl) scintillation crystals using modern Data Acquisition systems, which are designed on commercially available PCI electronics. The development of various techniques for optimization of the system’s performance will be discussed together with imaging results from test phantoms.

Speaker: Dr Efstathios Stiliaris (University of Athens & IASA)

15:00 → 16:30 Multi modality

15:00 Performance analysis of a low-cost small animal PET/SPECT scanner

PET and SPECT are nowadays fundamental techniques for the non-invasive monitorization of chemical pathways in living subjects, based on the emitted radiation of a radiolabeled pharmaceutical,, and are regarded as .powerful tools for the research with animal models of human diseases. However, imaging small rodents requirements in terms of resolution are not met by commercial human scanners and therefore several research groups world-wide have designed dedicated small animal scanners[1-5]. In this context, our goal is to specify a low-cost system capable of performing both PET and SPECT studies and adaptable to different geometries. In any case, it is obvious that the performance of the hybrid system must be comparable in terms of sensitivity, count rate and energy resolutions with existing state-of-the art devices of each modality. The aim of this work is to evaluate the expected performance at system level of a 4 detector head PET/SPECT scanner based on an acquisition front-end currently under development, considering a YAP/LSO phoswich detector. The performances estimations are obtained after the description and simulation of the proposed scanner with the Geant4 Application for Tomographic Emission (GATE) v2.2.0. The optical properties of the detector are optimized, in order to determine the optimum crystal finish Point sources and Derenzo phantoms are simulated and reconstructed for both modalities with the STIR library from and also with the ASPIRE software provided by the university of Michigan. The performance of the described setup, expressed in terms of image resolution and sensitivity, is 1.4mm/0.6% for PET (511 KeV) and 2.5mm/0.025% for SPECT (140KeV). These figures are compared with other existing scanners, with the conclusion that they are very close to state-of-the-art machines, despite the reduced number of detectors. These values will enable multimodal simultaneous acquisitions, providing a new insight into related metabolic processes, without sacrificing performance. [1] R. Lecomte, et al., "Design and engineering aspects of a high resolution positron tomograph for small animal imaging", IEEE Transactions on Nuclear Science, vol. 41, pp. 1446 - 1452, 1994. [2] S. Pavlopoulos and G. Tzanakos, "Design and performance evaluation of a high-resolution small animal positron tomograph", IEEE Transactions on Nuclear Science, vol. 43, pp. 3249 - 3255, 1996. [3] D. P. McElroy, et al., "First results from MADPET-II: a novel detector and readout system for high resolution small animal PET," at Nuclear Science Symposium Conference Record, vol. 3, pp. 2043 - 2047, 2003. [4] M. Streun, et al., "A PET system based on data processing of free-running sampled pulses," at Nuclear Science Symposium Conference Record, vol. 2, pp. 693 - 694, 2001. [5] J. Seidel, et al., "Resolution uniformity and sensitivity of the NIH ATLAS small animal PET scanner: Comparison to simulated LSO scanners without depth-of-interaction capability", IEEE Transactions on Nuclear Science, vol. 50, pp. 1347 - 1350, 2003.

Speaker: Mr Pedro Guerra (Universidad Politecnica de Madrid)

15:15 Preliminary Studies of a Simultaneous PET/MRI Scanner Based on the RatCAP Small Animal Tomograph

We are developing a scanner that will allow the simultaneous acquisition of high resolution anatomical data using Magnetic Resonance Imaging (MRI) and quantitative physiological data using Positron Emission Tomography (PET). The approach is based on the technology used for the RatCAP conscious small animal PET tomograph which utilizes block detectors consisting of pixelated arrays of LSO crystals read out with matching arrays of avalanche photodiodes (APDs) and a custom-designed ASIC. The version of this detector used for simultaneous PET/MRI imaging will be constructed out of all non-magnetic materials and will be situated inside the MRI field. We have demonstrated that the PET detector can be operated inside the MRI field using 511 keV gamma rays, and have obtained MRI images with its detector components located around various objects being imaged with MRI. The MRI images show minimal distortion in this configuration in which some components still contain traces of certain magnetic materials, and we plan to improve on the image quality in the future using completely non-magnetic components and by tuning the MRI pulse sequences. The combined result will be a highly compact, low mass PET scanner that can operate inside an MRI magnet without distorting the MRI image, and can be retrofitted into existing MRI instruments.

Speaker: Dr Craig Woody (Brookhaven National Lab)

15:30 Evaluation of Silicon Photomultipliers: A Promising New Detector for MR Compatible PET

The Silicon Photomultiplier (SiPM) is a novel, compact detector which shows great promise for use in MR compatible PET systems and other high resolution PET scanners for dedicated imaging applications. Like avalanche photodiodes (APDs), SiPMs are insensitive to magnetic fields, however they have the additional advantage of an extremely high intrinsic gain (about 1M), eliminating the need for preamplifiers and the accompanying shielding necessary for use in the bore of an MR scanner. They are also much faster than APDs, resulting in better coincidence timing resolution, and operate at a relatively low voltage (40 – 45 V). In addition, their compact size allows SiPMs to be used to extract depth of interaction information from scintillation crystals. A prototype detector is currently under investigation at the University of Toronto. The 1 mm x 1 mm active area of the SiPM consists of an array of 1440 individual avalanche photodiodes (microcells) operated in Geiger mode. Each microcell measures 20 microns x 30 microns, and all are manufactured on the same silicon substrate (pitch 25 microns x 35 microns) and connected in parallel through integrated quenching resistors. The output signal of the SiPM is thus the sum of the standardised signals from the individual microcells, and is proportional to the number of detected light photons, provided the number of light photons does not exceed the number of individual microcells. The detector prototype was operated without a preamplifier, and the output signal was sent directly to a shaping amplifier. Preliminary measurements made using 2 mm diameter LYSO crystals of varying length and an FDG source show encouraging results, from which a physical model of the next generation of multi-element SiPM detectors will be developed that should be well suited for use in a high performance MR compatible PET scanner.

Speaker: Dr David P McElroy (Sunnybrook and Women's College Health Science Centre, University of Toronto)

15:45 CT WITH A CMOS FLAT PANEL DETECTOR INTEGRATED ON THE YAP-(S)PET SCANNER FOR IN VIVO SMALL ANIMALS IMAGING

Introduction During last years the University of Ferrara has pursued an interest for in-vivo molecular and genomic imaging and has developed the YAPPET scanner [1], a integrated small animal PET-SPECT small animal scanner. The spatial resolution in PET is better than 1.8 mm whereas in SPECT it is better than 3.5 mm The PET sensitivity at the center is 640cps/µCi while in SPECT mode it is 1.1 cps/µCi with 4 detectors, constant over FOV. Now we are improving this device with a computed tomography (CT) in order to obtain simultaneously functional and morphological information and consequently we can, firstly, better visualize the region of interest and, secondly, obtain information on the attenuation coefficients necessary for improving reconstruction and quantitative measurements. At present we are characterizing and testing the CT-system with phantoms. Soon we will assemble the CT detector on the PET-SPECT device and at the congress in May we will present combined PET-CT and SPECT-CT images all acquired with a single scanner. Materials and Methods Materials: The CT system employed consist of a X ray tube and a large-area 2D image sensor. The tube has a tungsten anode with a focal spot of 0.6x0.6mm2; in the present measurements we used a 1.08mm aluminium filter giving an X-ray energy of 22 keV. The detector is a dual CMOS photodiode array (RadEye Inc.) whose large 49.2mm by 49.3mm active area consist of a 1024 by 1024 matrix of 48 µm silicon photodiodes. On the detector is deposited a uniform layer of Lanex Fast scintillato (130mg/cm2 of gadolinium oxysulfide Gd2O2S) which converts the X-rays into light. The detector is optimized to detect x-rays in 10keV to 50keV range. The flat panel detector, via an adapter board, is connected to the national instruments PCI-6111E board which acquires the signals. We have developed a dedicated LabVIEW programme that controls the image acquisition. In this way it is possible, via software, to change integration time, readout closck frequency and number of acquired images. The same LabVIEW program controls the X-ray exposure and motor motion and synchronizes all operations. In this preliminary stage we rotate the object under investigation keeping the detector and X ray tube fixed. The motor is a brushless DC-servomotors. In the final configuration the detector and the X ray tube will be fixed on the YAPPET gantry and consequently will rotate around the object under investigation. Methods: The X ray tube is 38cm from the object which in turn is at 4cm from the detector. The signals, that come from the CMOS detector, are converted with a 11 bits ADC and the counts per pixel are saved into hard disk in a raw-file format. Generally we use a clock at 625kHz, in this way the reading of the whole detector lasts 0.880 seconds. We have synchronized the X ray shoot and the acquisition in order to perform a reset before acquiring an image. Every image undergoes the following corrections: 1- dark current subtraction (at our integration time, 1sec, roughly 20 counts per pixel) 2- flat field “normalization” in order to correct the non uniform sensitivity pattern. Results and first CT-images We are aiming at high contrast images rather than high CT spatial resolution and the final voxel size will be of 0.5 x 0.5 x 0.5 mm3. CT images are taken with 180 projections over 180 degree in a total of 6 minutes. Finally we compared the attenuation coefficients obtained from the reconstructed images with literature values (XCOM: Photon Cross Sections Database-NIST Standard Reference Database 8) for 22keV X-rays (the mean energy of our spectrum). We report the values: µwater(literature at 22keV)=0.65cm-1;µPE(literature at 22keV)=0.349cm-1. µwater(measured)=0.581±0.015cm-1; µPE(measured)=0.361±0.006cm-1.

Speaker: Dr Giovanni Di Domenico (Universita' di Ferrara & INFN-Sezione di Ferrara - Italy)

16:00 Investigation of the LabPET Detector and Electronics for Photon-Counting CT Imaging

The development of new molecular probes targeting receptors with high specificity in selected cells and tissues highlights the importance of obtaining the anatomical context in Positron Emission Tomography (PET) imaging. This can be achieved using another imaging modality, such as X-ray Computed Tomography (CT) or Magnetic Resonance Imaging (MRI), but the anatomic and molecular images obtained sequentially with different scanners must subsequently be co-registered and are subjected to motion artifacts. Conventional CT imaging also contributes a significant dose, which may compromise the benefits of longitudinal molecular imaging studies in the same subject. To overcome these difficulties, we have investigated the use of the LabPET detector and electronics as a multi-modal detection system. Based on fast light emitting inorganic scintillators individually coupled to avalanche photodiodes and parallel, low-noise, fast digital processing electronics, the proposed detector front-end is suitable for coincidence detection of annihilation radiation (511 keV) in PET and for ultra-fast low-energy X-ray photon counting in CT. This combined detection system enables concurrent PET/CT imaging while achieving superior image contrast sensitivity for a given dose in CT photon-counting mode. Anatomical images with millimeter spatial resolution and sufficient tissue contrast for anatomical localization in small animals have been obtained with doses in the mGy range. The CT performance for dual-modality imaging of small animals was analyzed in terms of spatial resolution, noise and image contrast sensitivity as a function of dose.

Speaker: Mr Philippe BÉRARD (Université de Sherbrooke, Nuclear Medicine & Radiobiology)

16:15 Design of a Small Animal Multimodality Tomographer for X-Ray and Optical Coupling: Theory and experiments

A small animal multimodality tomographer dedicated to co-registration of fluorescence optical signal and X-rays measurements is under development in our laboratory. The purpose of such a system is to offer the possibility to get in vivo anatomical and functional information at once. Moreover, anatomical measurements will be used as a regularization factor in order to get the reconstructions of the biodistribution of fluorochromes more accurate and to speed up the treatment. The best configuration for X-rays imaging is obtained when the animal is illuminated perpendicularly and around its vertical axis. Furthermore, analytical solutions of the optical problem can be derived for simple geometries. The cylindrical geometry has been naturally chosen as a compromise: i) this geometry is suitable with X-rays cone beam CT of the whole animal body; ii) by immersing the animal in an index matching fluid, an analytical solution to the optical forward problem can be derived and introduced in a reconstruction scheme. An exact analytical solution to the Diffusion Approximation to the Radiative Transport equation used for modeling the optical forward problem has been established. For this geometry, we proceeded by analogy with the classical solutions to the equations of heat conduction in solids. The optical system is basically composed with a laser beam for excitation of the fluorochromes, and a CCD camera coupled with a chromatic filter for the fluorescence detection. Experimental measurements on cylindrical phantoms with different fluorescent inclusions geometries have been performed as a validation. The cylindrical analytical approach has been compared to a Finite Element Method approach: analytical approach brings a significant gain of computing time.

Speaker: Dr anabela da silva (CEA-LETI)

16:30 → 17:00 Coffea break

17:00 → 18:30 Advances in imaging Technology

17:00 FURTHER DEVELOPMENT OF SCINTILLATION MATERIALS FOR PET SCANNERS

The growing demand on PET methodology for a variety of applications ranging from clinical use to fundamental studies triggers research and development of PET scanners providing better spatial resolution and sensitivity. These efforts are primarily focused on the development of advanced PET detector solutions and on the developments of new scintillation materials as well. However in spite of tremendous efforts of the researchers the new instrumentation developments are mainly based on Lu containing scintillation materials introduced in the last century. LSO, LYSO, LuAP, LuYAP crystals still remain the best PET scintillators in spite of the recent developments of bright, fast but relatively low density bromide scintillators [1]. At the same time Lu based materials have several drawbacks which are high temperature of crystallization and relatively high cost compared to alkali-halide scintillation materials. Here we describe recent results in the development of new scintillation materials for PET application. In the research reported we aimed to achieve the following goals: 1) temperature decrease of the crystalline materials synthesis; 2) development of high Z complex compounds without Lu or with its partial replacement; 3) development of crystal growth techniques for shaped crystals to minimize losses of the material at the mechanical treatment. 1. International Conference on Inorganic Scintillators and their Industrial Application SCINT2005. Book of Abstracts, Alushta, Ukraine, September 19-23, 2005

Speaker: Prof. Mikhail Korjik (Institute of Nuclear Problems, Minsk, Belarus)

17:15 Silicon Drift Photodetectors for scintillation readout in medical imaging

In the attempt to improve system performances of position sensitive gamma-ray detectors with contemporary ease of use and compactness, the use of silicon photodetectors for scintillation readout has become of increasing interest. With respect to the long-established photomultiplier tubes (PMTs), silicon photodiodes (PDs) have the advantages of higher quantum efficiency, smaller dimensions and lower biasing voltages. Unfortunately, the absence of a mechanism of electronics multiplication as in PMTs makes conventional PDs limited in their scintillation readout performances, because of the very high noise contribution to the overall resolution. Avalanche photodiodes (APDs) combine the high quantum efficiency and compactness of PDs with the benefit of a moderate avalanche multiplication gain. However, the statistical component is again affected by the statistics of the multiplication itself (noise factor), and the noise component is still appreciable mainly in the high energy range. Moreover, the sensitivity of the gain to temperature and biasing variations represents a potential practical drawback in the use of APDs arrays for the application. As an alternative to the mentioned photodetectors, silicon drift detectors (SDDs) have recently shown to achieve excellent performances in scintillation light detection. The SDD is a photodetector characterized by an area-independent, low value output capacitance (~100 fF), which is one of the major factors limiting the noise performances of the other silicon detectors’ technologies. SDDs used for CsI(Tl) scintillation readout have already shown to achieve state-of- the-art energy resolution in g-ray spectroscopy, and monolithic arrays of SDDs have been recently experimented for the development of Anger Cameras for high-resolution gamma-ray imaging. In a first prototype of a small Anger Camera, a SDD array of 1cm2 active area has been coupled to a single CsI(Tl) scintillator. An intrinsic spatial resolution better than 200 um FWHM with 122 keV photons has been measured with this prototype. Aim of this work is to present the results of the experimental characterization of a high-resolution LaBr3:Ce scintillator coupled to a single SDD photodetector. The active area of the SDD is of 30 mm2 and antireflective coatings have been implemented on its entrance window. Excellent energy resolutions were measured at room temperature, thanks to the low leakage current of the detector: 2.7% at the 137Cs 661.7 KeV line and 6.1% at the 57Co 122 KeV line. A resolution of 5.7% at 122 KeV line was measured at 0°C. In addition, a new large-area monolithic array of 77 SDDs, developed in the framework of the project DRAGO-INFN and supported by Italian INFN, is also presented. The goal of the DRAGO-INFN project is to develop a gamma-ray imager characterized by a submillimeter position resolution, to be used for dedicated imaging in medicine and for small animal imaging. The detector, with its active area of about 6.7 cm2 , is the largest monolithic array of SDDs with on-chip JFETs yet produced for the mentioned applications.

Speaker: Carlo Fiorini (Politecnico di Milano and INFN Italy)

17:30 Silicon Photomultipliers and their bio-medical applications

A brief overview of the history of the development of the so-called 'Silicon Photomultipliers' (SiPM), more precisely - multi-pixel avalanche photodiodes operating above breakdown (or in 'Geiger' mode) will be given. Main features, which determine their potential for bio-medical applications, will be described. Some examples of particular on-going or future developments in this field will be given.

Speaker: Eugene Grigoriev (Institute for Theoretical and Experimental Physics (ITEP))

17:45 Development of a novel geometrical concept of a high

The concept was proposed as a competitive and innovative approach, mainly for a high resolution brain PET scanner, allowing a parallax free 3D image reconstruction with Compton enhanced sensitivity. It is based on axially oriented matrices of 16 x 13 long polished LYSO scintillator bars (e.g. 3.2 x 3.2 x 150 mm3) optically coupled at both extremities to segmented Hybrid Photon Detectors (HPD) readout by mean of a VLSI auto-triggering fast Front End Electronic (FEE) encapsulated in the detector body. The status of the development program will be discussed after an introduction in order to recall the basic principles of the concept and the subsequent requirements to the readout system. The progress on the characterization of the crystal bars, a key component of the project, will be presented in a separated contribution. The design of the HPDs and of the FE readout electronic will be described and the performances obtained with a prototype PET-HPD reported. Finally, the design of an Event Driven Read-out System under development will be presented.

Speaker: Peter Weilhammer (INFN Perugia)

18:00 Nuclear medical imaging using beta+ gamma coincidences from 44Sc radio-nuclide with liquid xenon as detection medium

We report on a new imaging technique based on the measurement of the emitter location in the three dimensions with a few mm spatial precision using beta+ gamma emitters. Such coincidences arise from a new kind of radio-nuclides, potentially produced at the ARRONAX cyclotron of Nantes, which emits a gamma-ray quasi simultaneously with the beta+ decay. The principle is to reconstruct the intersection of the classical line of response (obtained with a classical PET-camera) with the third gamma-ray. The emission angle measurement of this additional gamma-ray involves the use of a Compton telescope for which a new generation of camera based on a liquid xenon time projection chamber is considered. Geant4 simulations of such a large acceptance liquid xenon Compton telescope camera combined with a commercial microPET (LSO crystals) have been performed and the obtained results will be presented. It demonstrates that a direct location of the 44Sc emitter with 2.5 mm spatial resolution in the 3 dimensions is achievable in counting mode.

Speaker: Mr Cyril Grignon (Subatech, EMN-IN2P3/CNRS-Universite)

19:00 → 23:00 Conference Dinner

Friday 12 May

08:30 → 13:15 Molecular Imaging

08:30 Combining Morphology and Molecular Imaging

From autoradiography to planar X-rays, Computed Tomography (CT) and Magnetic Resonance (MR), morphology and structure has been the mainstay of biological and medical imaging for over a century. While structural changes may suggest the presence of disease, functional changes are more sensitive indicators of early- stage pathology, and where cancer is concerned, early detection is the key to a favorable prognosis. Since molecular imaging offers the potential to quantitatively image functional changes in vivo, it is assuming an increasingly important role in the identification, staging and re-staging of human disease. Specifically, Positron Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) are sensitive techniques to map human physiology non- invasively through the use of high-resolution imaging devices and appropriate radioactively-labeled biomarkers. However, such metabolic maps do not offer the structural detail associated with anatomical imaging techniques such as CT and MR and therefore dual modality devices such as PET/CT, SPECT/CT or PET/MR that combine both structural and functional information offer a more complete and accurate assessment of the status of disease. Within the past five years, the introduction and rapid adoption of dual modality imaging technology has significantly impacted the medical imaging field. Building on these developments for human imaging, microCT, microSPECT and microPET scanners with sub-millimeter resolution have essentially replaced conventional autoradiography for studies of disease models involving small and medium-sized animals. This presentation will assess the role of multimodality instrumentation within the context of translational research from animal models to human disease, with particular emphasis on cancer. Some recent developments and future directions of multi-modality imaging technology will be highlighted.

Speaker: Prof. David Townsend (University of Tennessee Graduate School of Medicine)

Slides Fri0830-P1.pdf Fri0830-P1.ppt

09:00 Reevaluation of MR Spectroscopy for the Coming Molecular Imaging

Regarding that new molecular imaging technologies mostly emerge from isotopic imaging, it is important to reevaluate preceding technologies of other modalities and their current role in clinical medicine in order to understand what we presently know and what are needed. Among many modalities, MR spectroscopy (MRS) should be reemphasized because it is a robust in-vivo technique which can detect intrinsic metabolite molecules in living tissues and evaluate their concentrations. Proton (1H) and phosphorus (31P) MRS are applied to patients using clinical high-field (higher than or equal to 1.5T) MR units. Detectable metabolites by MRS are those related to the energy and membrane metabolism, organ-specific functions, or fermentation processes, such as ATP, (phospho)creatine, lactate, (phospho)choline, (phospho)ethanolamine, n-acetyl aspartate in the brain, citrate in the prostate, amino acids, lipids, etc. Metabolite signals are obtained from the defined region-of-interest in the tissue using a sophisticated signal localization MR sequence. Measurement time is usually several minutes in 1H MRS and 30 minutes in 31P MRS for a patient. Analysis of signal intensities of metabolites in MR spectra is helpful for diagnosis of diseases, like tumors, inflammation, infection and degenerative diseases. Especially, treatment follow-up of patients is the another area of application of MRS. Viability of cancer cells can be estimated based on signal intensities of ATP and phosphocholine/phosphoethanolamine in a 31P MR spectrum; the former is a measure of energy activity and the latter membrane turnover or cell proliferation rate. Therefore, MRS provides a biochemical evaluation of treatment effect of cancers, which otherwise can hardly be known. Although reduction in tumor size is accepted as an indicator of an effective treatment, discrepancy is sometimes observed between morphological evaluation and biochemical one; MRS suggests still a high viability of cells in the reduced residual tumor. Further treatment or change of therapeutic regimen are needed for those cases. Our experience will be presented. As conclusions, reevaluation of MRS might accurately present needs for the emerging new molecular imaging because MRS is the only available technique which provides in-vivo biochemical information in clinical medicine.

Speaker: Dr Keiko Imamura (Department of Radiology, St. Marianna University School of Medicine)

09:15 An easy to use imaging tool and radiopharmaceutical agent derived from CCK4 for internal radiotherapy : synthesis and assessment of an original biovector

Introduction Cancer constitutes one of the first causes of mortality in the world. This proliferative disease is characterised by two evolutionary components made of regional extension and metastatic diffusion. The therapeutic success of cancer treatment depends on precocity of the diagnosis, of extension assessment and initial treatment of primitive tumour and its metastatic dissemination. The present work is part of the French National Cancer Fighting Plan as illustrated in south France with the founding of the “Cancéropôle”. For tumoral targeting, biomolecule vectors complexed with two different isotopes (gamma- and beta-emitters) seem to be a suitable strategy. Indeed, technetium-99m can be used for diagnosis and rhenium- 186/188 for radiotherapy and rhenium-185/187 for chemotherapy [1]. The tetragastrin (CCK4), an effector related to gastrin, exhibits specific affinity for CCKb receptors which are preferentially expressed by some tumoral cells in digestive tract. The aims of this work are (i) synthesis of a beta-ala-CCK4 biovector coupled to SN3 ligand and (ii)development of technetium and rhenium complexation method using physiological pH and temperature conditions and (iii) pharmacological in vitro and in vivo studies of these labelled vectors. Methods S-(1-ethoxyethyl)mercaptoacetyl-triglycyl-b-alanine-CCK4 vector was obtained by successive condensations of activated amino acids on resin and characterised with HPLC-UV, 1H-NMR. The labelled biovectors (185/187Rhenium and 99/99mTechnetium) were (i) incubated with CCKB cell receptors and their affinity was tested by tetragastrin competition and (ii) injected into healthy Wistar rats and nude mice with pancreatic adenocarcinoma grafts (AR4-J). Pharmacokinetics and tissue distribution were monitored by NaI gamma camera imaging. Results The biovectors were obtained with high chemical purity and were efficiently labelled (>95%) in one step under pH 7-8 and 45°C conditions. These labelled biovectors present remarkable pharmacokinetic properties related to low background noise resulting from lack of hepatic uptake. Moreover, labelled CCK4 has a good affinity for CCKB receptor (IC50 8. 10-8M). This result allows to conclude that linking the SN3 chelating agent to CCK4 does not alter the conformation and biochemical properties of this peptide and then may be proposed as radiopharmaceutical kit in routine cancer treatment. 1- Belhadj-Tahar H. et Darbieu M.H. et coll. Med. Nucl. Imag. Fonct. Metab. 2004, 28 (3) : 101-9

Speaker: Dr Hafid BELHADJ-TAHAR (Groupe Santé Recherche)

09:30 In vivo PET evaluation in tumour bearing rats of [18F]-2-fluoromethyl-L-phenylalanine as a new potential tracer for molecular imaging of brain and extra-cranial tumours in humans with PET.In vivo PET evaluation in tumour bearing rats of [18F]-2-fluoromethyl-L-phenylalanine as a new potential tracer for molecular imaging of brain and extra-cranial tumours in humans with PET.

The Na+-independent L-type LAT amino acid transport system 1 for neutral and lipophilic amino acids has been shown to be increased in tumour tissue relative to normal tissue, and the LAT system has been regarded as a key-point for the development of new amino acid based tumour tracers for molecular imaging. We have proven in vitro and in vivo that the new compound 2-I-L-phenylalanine is taken up in R1M and in several types of human cancer cells by LAT for the major part. This radio-iodinated amino acid show a very high tumour selectivity when compared to the work horse for oncology PET, [18F]-FDG, which is taken up considerably in brain and inflammatory tissue. Therefore we developed a new fluorinated phenylalanine analogue, [18F]-2-fluoromethyl-L-phenylalanine ([18F]-2-Me-L-Phe), considering that the special volume of FCH3 is comparable with that of the I atom in 2-I-L- phenylalanine and which could be prepared with the ease of [18F]-FDG to allow clinical routine. The substrate molecule for radio labeling 2-Bromomethyl-L-phenylalaine was custom prepared by radical bromination of Me-L-Phe. [18F]- for bromo exchange is performed within 5 minutes in conditions comparable to FDG synthesis with a radiochemical yield of at least 85%. After deprotection (15 minutes) the [18F]-2-fluoromethyl-L- phenylalanine is recovered n.c.a. with a high purity by means of semi-prep HPLC in a solution which after make-up and sterilization is ready for injection. 3.7 – 10 MBq were injected into R1M rat rhabdomyosarcoma bearing tumours with a volume ranging from 1cm3 to 2.7 cm3. Imaging was performed with a human Siemens Acsel PET camera from 5 to 45 minutes p.i.. Quantification of activity uptake occurred by conventional algorithms. The tumour (shoulder or flank) / background (contra lateral shoulder or flank) and tumour / blood ratios obtained from PET acquisition were at least 3. This was confirmed by measurement of the activity in organs and tissue of interest after dissection. A small tumour implanted near a kidney could be well visualized completely separated from this kidney. Moreover in all tumours the “living” tumour cells can clearly be differentiated from necrotic tissue. This proves that [18F] 2-fluoromethyl-L-phenylalanine has a great potential as a new tracer for specific tumour diagnosis with PET.

Speaker: Prof. John Mertens (BEFY, Vrije Universiteit Brussel)

09:45 Monitoring of Tumor Growth and Metastasis Potential in MDA-MB-435s-tk/luc Human Breast Cancer Xenografts

Recently, molecular imaging of reporter gene expression provides a rapid, sensitive, and noninvasive monitoring of tumor behaviors. In this study, we reported the establishment of novel animal models using immunodeficient mice for longitudinal examination of tumor growth kinetics and metastatic spreading in vivo. The highly metastatic MDA-MB-435s cell line was engineered to stably expressed herpes simplex virus type 1 thymidine kinase (HSV-1-tk) and luciferase. Both 131I- FIAU and D-luciferin were used as probes. For orthotopic tumor formation, MDA-MB- 435s-tk/luc cells were implanted into the first nipple of 6-week-old female NOD/SCID mice. For metastatic study, cells were injected via the lateral tail vein. Mice bearing MDA-MB-435s-tk/luc tumors were imaged for tumor growth and metastatic sites using Xenogen IVIS50 bioluminescent system. Gamma scintigraphy, whole-body autoradiography were also applied to confirm the localization. The result of bioluminescent imaging showed that distinct photon accumulation could be detected and quantified in the femur, spine, spleen, salivary gland, lymph nodes, and muscle at 16 weeks post i.v. injection. Histopathological examination further verified tumor metastasis at those organs as well. We concluded that establishment of MDA-MB- 435s-tk/luc human breast carcinoma-bearing mouse model combined with multimodalities of molecular imaging can facilitate studies on the molecular mechanisms behind invasion and metastasis, and also allow a more convenient, sensitive and noninvasive imaging protocol for the diagnosis of the diseases as well as for the evaluation of therapeutic efficacy of newly developed drugs or new treatment trial in preclinical studies.

Speaker: Ms Ya-Fang Chang (Institute of Radiological Sciences, National Yang-Ming University, Taipei, Taiwan)

10:00 Preclinical evaluation of new radioligands of cholecystokinin/gastrin receptors in endocrine tumours xenograft nude mice.

Aim: Whereas somatostatin receptor scintigraphy has been proven a valuable tool for staging gastrointestinal endocrine tumors, its sensitivity and accuracy in other neoplasm, such as metastatic medullary thyroid cancer (MTC) or small cell lung cancer (SCLC), is limited by the fact that the somatostatin receptors are not expressed in all these tumours and metastasis or that the expression of these receptors can change during the evolution of the pathology. The CCK-2 (cholecystokinin)/gastrin receptors (RCCK2) are overexpressed in up to 90% of MTC and 60% in SCLC but not in corresponding healthy tissues. Thus, they represent an ideal target for the diagnosis and internal radiotherapy of these tumours. The objective of our study was to compare previously published CCK radioligands (111In- DTPA-CCK8) to newly synthesized ligands of the RCCK2 containing new generations of metal chelating in order to increase the efficacy of the RCCK2 targeting and to limit nephrotoxicity. Materials and Methods: Derivatives of CCK8, CCK4 and gastrin peptides have been developed and were covalently coupled to the new synthetized chelating agents to be further labelled with 111Indium. The stability and affinity of the radiolabelled peptide were studied in vitro and in vivo. Nude mice, bearing tumors from the human SCLC cell line NCI-H69 and the human MTC cell line TT, were intravenously injected with 10 MBq of radiolabelled peptide. In vivo scintigraphy of these various peptides were performed 24 h post injection : a dynamic acquisition of planar images was performed using a gamma camera (Millennium VG-GEHC) equipped with low energy high resolution collimators. Images analysis was done on a Xeleris workstation (GEHC-Waukesha). Thereafter biodistribution studies (%ID/g tissue ) were done with a gamma counter (Wallac Wizard). Results: We have elaborated and optimised the conditions of radiolabelling leading to a radiochemical purity > 85%. In vitro studies confirmed that the CCK compounds tested displayed a high affinity for the RCCK2 (nanomolar range). Scintigraphic studies of xenograft mice showed significant tumour uptake with a high target to non target ratio. Ex vivo studies confirmed the efficiency of RCCK2 targeting with these original radiolabelled CCK analogs. A biodistribution study showed physiological distribution (stomach, pancreas), renal excretion (gallbladder, kidneys) and tumor uptake at 24 h post injection. Conclusion: According to these studies, these new radiolabelled compounds seem to have promised stability and high receptor affinity thus these “optimised” CCK ligands coupled to specific chelators might be potential new candidates for molecular imaging and internal radiotherapy of tumours overexpressing the RCCK2 receptors.

Speaker: Dr SEVERINE BRILLOUET (INSTITUT CLAUDIUS REGAUD)

10:15

10:30 Coffea break

11:00 Luminescent probes for optical in vivo imaging

Molecular imaging allows a better insight of biological mechanisms in vivo, such as the follow-up of gene expression, drugs biodistributions, the assessment of therapies, and can be used for a wide variety of applications, such as cancer, cardiovascular diseases, inflammation… If several techniques such as nuclear medicine (SPECT, TEP), molecular resonance imaging (MRI) and X ray computed tomography (CT) already bring information in vivo, the optical range of interactions between light and tissues has started to be fully exploited more recently. The advantages of the optical methods lie in their low cost, low manipulation constraints (no radioactivity), short acquisition times and high sensitivity. As new optical imaging techniques such as bioluminescence and fluorescence tomography emerge and appear as new modalities to assess biological events in small animals, the need for suitable optical probes arises. We will present new molecular probes for fluorescence in vivo imaging of tumours in mice. The core of these probes is constituted by a cyclodecapeptide vector, the RAFT molecule, which has a nearly planar conformation. Both a luminescent reporter for imaging and eventually a drug can be grafted on one face of the RAFT, while the other can be independently functionalized by four biological ligands for tumour targeting1. We chose to use the cRGD peptide that allows targeting of the V3 integrin receptors of the endothelium tumour cells. The imaging function can be a classical fluorescent dye, such as Cy5, or a more sophisticated activatable fluorescent function. For these activatable probes, the fluorescence signal is inhibited until the probe has been internalized in the targeted cells. Image contrast is thus dramatically improved. Examples of specific in vivo imaging of internal tumours in nude mice bearing IGROV1 metastatic nodes (human ovarian cancer) will be presented using the different classical and activatable fluorescent probes as contrast agents. 1. D. Boturyn, J.L. Coll, E. Garanger, M. C. Favrot, P. Dumy, J. Am. Chem. Soc. 2004, 126(18), 5730-5739.

Speaker: Dr Philippe Rizo

11:15 Molecular imaging with radionuclides: a powerful means for studying biological processes "in vivo" process

Recently there has been a growing interest in molecular imaging techniques by radionuclides. In the imaging techniques panorama incluidng MRI, x-ray CT, optical, etc., they have a specific role due to their unique features. For example, they are extremely sensitive (picomolar) that is needed for imaging biological processes “in vivo”. A wide range of human diseases can be studied in animal models. Also, it is possible to provide early detection of small tumors with high specificity. Nevertheless, the techniques are technologically challenging because of the concurrent requirements in both high spatial resolution and high sensitivity. The limitation of the sensitivity due to the collimation is well known and affects the performance of detector systems, especially if only radiopharmaceuticals with limited uptake are available. The “electronic” collimation technique used in Positron Emisson Tomography (PET) has intrinsic limitations in terms of spatial resolution. Moreover, it is often not well suited for some applications, for example involving protracted longitudinal measurements, or for the reasons of applicable biochemistry. An electronic collimation method for single gamma emissions using Compton camera techniques is rather complicated, expensive and not yet fully developed. Multi-pinhole and coded aperture collimation are promising solutions to improve sensitivity. This is the case at least for “small volumes” imaging involving small animals. In this paper we will present simulations performed for optimizing the performances of dedicated detectors with multi-pinhole and coded aperture collimation taking into account factors that affect spatial resolution, dimension of scintillator pixels, and pixel identification, number of pixels per unit area, signal-to-noise (SNR) and contrast. Phantom as well as preliminary small animal imaging measurements have been performed to evaluate the performance of the molecular imaging techniques for specific small animal imaging studies.

Speaker: Dr Evaristo Cisbani (Istituto Superiore di Sanita' - Istituto Nazionale di Fisica Nucleare)

11:30 Accuracy of partial volume effect correction in clinical molecular imaging of dopamine transporter using SPECT

Objectives: Partial volume effect (PVE) is a major source of bias in brain SPECT molecular imaging of dopamine transporter: radioactivity concentration measured within striatal volumes can be underestimated by more than 50% because of PVE only. Various PVE corrections appropriate for SPECT and making use of anatomical data have been developed. These methods yield encouraging results in studies involving simulations and/or phantom acquisitions. However, their accuracy in clinical data is difficult to demonstrate, first because the accurate value (gold standard – GS) of the parameters to be estimated is usually unknown, and second because PVE correction is greatly affected by the accuracy of segmentation of anatomical data and of co-registration of SPECT with anatomical images. The objective of this study was to assess the accuracy of a PVE correction on clinical dopamine transporter SPECT studies. Method: 23 patients underwent MRI and dopaminergic neurotransmission 123I-FP-CIT SPECT. MRI studies consisted in 3D T1-weighted sequences with 2 mm thick slices. SPECT acquisitions were performed 4 hours after the injection of 185 MBq of 123I--FP-CIT. A 3 headed Prism 3000 XP camera equipped with low energy ultra high resolution fan beam collimators and with a transmission source device was used. SPECT projections were corrected for scatter and reconstructed using OSEM (12 subsets, 12 iterations) including attenuation compensation. Striata were manually segmented on the MR images. The binding potential values (BP) were measured in the striatal volumes of interest after coregistration of MRI data to SPECT data by mutual information maximization. BP were defined as [S-NS]/NS, where S was the striatal activity and NS represented non-specific activity measured in a posterior volume of interest. These values were calculated without and with an original PVE correction previously described (Soret et al, Eur J Nucl Med Mol Imaging 2006). In addition, for each patient, a Monte Carlo simulation of the patient SPECT scan was performed using the SimSET code. The anatomical data needed for each simulation were derived from the segmented MRI of the patient, and 123I activity concentrations were set equal to those found from the patient SPECT data after all corrections. The simulated projections were processed exactly as the real data. For the Monte Carlo simulations where true simulated BP values were known, percent biases in BP estimates were calculated. For the real data, an evaluation method that did not require the GS to be known was used. This method, derived from the one by Hoppin et al (IEEE Trans Med Imaging 2002), simultaneously estimates the GS and a quadratic relationship between the observed and the GS values, assuming the GS follows a beta law. It yields a surrogate mean square error (sMSE) between the estimated values and the estimated GS value. Results: The averaged percent difference between BP measured for real patients and BP measured on the corresponding simulated patients was 0.7±9.7% without PVE correction and was -8.5±14.5% with PVE correction, suggesting that the simulated data reproduced the real data well enough. For the simulated patients, BP was underestimated by 66.6±9.3% on average without PVE correction and overestimated by 11.3±9.5% with PVE correction, demonstrating the greatest accuracy of BP estimates when PVE correction was used. For the simulated data, sMSE obtained by assuming the GS was unknown were 27.3 without PVE correction and 0.90 with PVE correction, confirming that our sMSE index properly captured the greatest accuracy of BP estimates with PVE correction when the GS was not supposed to be known. When considering the real patient data for which no direct bias calculation was possible as the GS was unknown, sMSE in BP estimates was 50.8 without PVE correction and 3.5 with PVE correction. These results were very consistent with those obtained on the simulated data, suggesting that for clinical data, and despite probable segmentation and registration errors, BP were more accurately estimated with PVE correction than without. Conclusion: By simulating real patient data and using an evaluation method appropriate for assessing the accuracy of estimation methods when the GS is unknown, we gathered evidences that the PVE correction considered in this study was very efficient in real patient to greatly reduce the error in BP estimates in clinical molecular imaging of dopamine transporter using SPECT.

Speaker: Dr Marine Soret (U678 Inserm UPMC, HIA Val de Grace)

11:45 New radiopharmaceuticals based on peptides and antibodies: Planar and SPECT small animal imaging with high resolution gamma cameras.

Aim. New radiopharmaceuticals are growing attention for cancer diagnosis and therapy. This work deals with the use of an updated High resolution (HR) camera for small animal imaging. Matherials and Methods. We studied three radiopharmaceuticals: two antibodies (MoAb) and one peptide (P) labelled with 99mTechnetium in 20 rodents. Balb/c nude mice xenografted with cancer cells or T- lymphocytes (T-LY) were injected with 3.7 MBq of MoAbs. A series of three mice were xenografted respectively with 5, 10 and 20 x 106 T-LY. 99mTchmetium were studied with Anger camera (AC) fitted with parallel hole, high resolution, low energy collimator and our new HR device. Five rats, weighting 250g each, were tested with peptide and images were acquired only with HR SPECT. Our HR camera (Li-Tech, Italy) is based on multi-crystal and Hamamatsu position sensitive phototubes technique (PSPMT) with spatial resolution of 2 mm. It was equipped with a mechanical arm to scan the entire body of rodents or to perform SPECT. HR camera, collimators and crystal-collimator-PSPMT matching are covered by 5 US and EU patents. 99mTc-Mo-Abs were directed against a cancer specific protein (MoAb 1) and T lymphocytes activated versus ulcerative cholitis (MoAb 2) and were studied with whole body planar scan 6, 12 and 24 h after injection. The 99mTc-P was studied with SPECT 180°, 36 views, from 1 to 3 h after injection. SPECT images were reconstructed with a rotation-based version of OSEM. Results. Whole body scan of nude mice required 3 ±0.5 min at 6h, 10± 3 min at 12 and 20±9 min at 24h when acquired with the 1 inch2 FOV HR camera. These times were not significantly different with Anger camera. The 4 inch2 FOV HR camera performed scans respectively in 1± 0.2 (P<0.01), 4± 2 ( P< 0.01) and 9± 5 (P<0.01) min. SPECTs of rat brain were acquired in 10±3 min. Both Anger camera and HR cameras imaged the xenografted tumours. However necrotic zones inside tumours were shown only by HR cameras. All X of T-LY were shown by HR cameras with direct relationship among number of xenografted cells and Target/Background ratio, whereas Anger camera showed only xenografted of 20x106 T- LY. HR SPECT showed Amygdale, Hippocampus, and dorso-lateral nuclei of thalamus in rat brain. All these sub-cerebral structures of rat were measured at autopsy as slightly larger of 2 mm on mayor axis. Conclusion. HR small animal imaging is useful in the study of new radiopharmaceuticals. Our device is faster than Anger camera and is able to show uptake sites missed by Anger camera

Speaker: Prof. Francesco Scopinaro (Univ. Roma "La Sapienza" - Italy)

12:00 Coherent Anti-Stokes Raman Scattering (CARS) Microscopy

Key words: Four-wave mixing, Coherent anti-Stokes Raman scattering (CARS) Microscopy, chemical imaging Coherent anti-Stokes Raman scattering microscopy (CARS) is a new approach for chemical imaging of molecular systems, with high sensitivity, high spatial resolution, and three dimensional sectioning capability, without using fluorophores that are prone to photobleaching. This technique permits to map selectively molecular species, by using vibrational properties of their chemical bounds. CARS is described by a four wave-mixing process, where the signal intensity depends nonlinearly on the incident intensities, and generated in a direction determined by the phase-matching condition. The CARS signal can be detected in both forward and backward directions [1]. The two signals provide complementary information about a sample: forward-detected CARS (F-CARS) microscopy is suitable for imaging objects of a size comparable to or larger than the excitation wavelength. Epi-detected CARS (E- CARS) microscopy provides a sensitive means of imaging objects having an axial length much smaller than the excitation wavelength because it avoids the large backgrounds from the solvent [2, 3, 4]. In this presentation, theoretical and experimental aspects of CARS microscopy in collinear excitation beam geometry are shown. Particular attention is given to the underlying physical principles behind the new features of CARS signal generation under tight focusing conditions. A brief overview of the instrumentation of CARS microscopy and its experimental characterization is provided by means of imaging of model systems and live unstained cells. References [1] A. Volkmer, J-X. Cheng, X.S. Xie, phys. Rev. Lett. 87, 23901 (2001). [2] J-X. Cheng, X.S. Xie, J. Phys. Chem B 108, 827 (2004). [3] J-X. Cheng, A. Volkmer, X.S. Xie, J. Opt. Soc Am. B 19, 1363 (2002). [4] N. Djaker, PF. Lenne, H. Rigneault, Proc. SPIE. 5463, 133 (2004).

Speaker: Mrs NADIA DJAKER (INSTITUT FRESNEL)

14:00 → 15:30 Perspectives

14:00 Contribution of HEP Electronics Techniques to the Medical Imaging Field

The purpose of this study is to show how innovative concepts of compact, pipelined, deadtimeless and ”Time Of Flight” capable electronics similar to those developped for High Energy Physics experiments (LHC and post LHC) could be fairly and easily transferred to the medical imaging field through clinical Positron Emission Tomography scanners. The two overriding weaknesses of PET camera readout electronics, namely timing resolution and dead-time, were investigated analytically and by Monte-Carlo simulation. Results shows that there is rather space available for count rate enhancement, especially through a huge decrease of the timing resolution well below the nanosecond. An optimized solution using basic building blocks of HEP for a generic read out electronic chain is proposed and discussed.

Speaker: Mr Pierre Etienne Vert (LPC Clermont and DAPNIA)

14:15 Towards a Continuous Crystal APD-based PET Detector Design

Monolithic scintillators are an attractive alternative to pixelated detector schemes in the design of high- sensitivity, high resolution PET systems. The shape of the scintillation light distribution measured by a position sensitive photo detector contains (x,y,z) information on the interaction point in the crystal. The absence of inter-crystal optical reflectors significantly increases the sensitivity of such a detector. In addition, the size of the continuous scintillator block can be larger than the sensitive area of the photo detector used to read it out. This results in a detector design with very high packing fraction without the need for more complex solutions such as coupling crystal pixels to PSPMTs using optical fibers. To test this approach of photon localization in PET, we have examined the performance of prototype detectors, based on continuous 20x10x10 or 20x10x20 mm3 LSO blocks and read out by 1 or 2 Hamamatsu S8550 APD arrays. Cramer Rao lower bound estimation on the spatial resolution,based on simulation data, were less than 0.9 mm FWHM for interactions occurring in 90% of the crystal volume. The lower bound increases for interaction occurring further away from the APD or closer the to crystal edge. The simulation results also allowed us to evaluate the performance of various positioning algorithms (e.g. neural networks, support vector machines) and their training procedures. The advantage of these positioning algorithms is the inherent DOI correction when they are trained for events impinging under a given angle. Furthermore, measurement on experimental setups showed that an average intrinsic spatial resolution of 1.5 +- 0.4 mm can be achieved. A linear regression of the average measured position versus the true incidence position showed a slope of 1.03 (r=0.999) to within 2 mm of the crystal’s edge. The energy resolution is 11-13 % FWHM and a time resolution of 2.5 ns was achieved. To further maximize the sensitivity, the possibility of using trapezoidal LSO blocks was investigated, both through simulation and experimental measurements. No significant difference in spatial and energy resolution was observed in comparison to the rectangular LSO blocks. Finally a prototype demonstrator using two detector modules in a rotating configuration was build to demonstrate the expected performance in a PET system. Images of point sources obtained with this device showed a resolution in the reconstructed image better than 2 mm FWHM.

Speaker: Dr Peter bruyndonckx (Vrije Universiteit Brussel)

14:30 High spatial and energy resolution lanthanum bromide SPET scanner for molecular imaging

Last year many research groups have been attracted from the first samples of LaBr3:Ce crystals available from St.Gobain. Different applications are under investigation in particular in the field of medical imaging. PET, SPECT and Compton camera instrumentation seem to take advantage from the performances of this new scintillation crystal. In fact, it shows many advantages with respect to previous scintillators: excellent energy resolution, (3% and 6% at 662 keV and 140 keV photon energy respectively)very good radiation absorption properties, high photofraction (similar to NaI(Tl)) and high response speed (16 ns of scintillation decay time. Although St.Gobain is able to produce volumes as large as 3 inch diameter by 3 inch thick, some limitations are arising from high hygroscopicity and material fragility that introduce serious concerns in pixelated manufacturing. At present time, LaBr3:Ce crystals are available with continuous shape covering 5x5 cm2 and thickness up to one inch. In 2004, trusting on election properties for SPECT application of lanthanum trihalides scintillators cerium-doped, INFN promoted the development of a scintillation camera based on continuous crystal shape. During the same year, the preliminary results have been obtained with a small LaCl3:Ce one inch square. In 2005, with the availability of LaBr3:Ce crystal, we realized a small 5x5 cm2 camera by integral assembling a continuous crystal with an Hamamatsu Flat panel PMT H8500. Such tube offers a superior imaging performance with respect to previous generation ones. The tube with, 64 anode matrix, when coupled to a continuous crystal well simulate a miniature Anger Camera. Taking into account the new scintillation properties offered by LaBr we revisited the potentiality of position arithmetic algorithm analysed 30 years ago. For the statistics nature of spatial resolution when position is carried out from the scintillation light centroide, a strong improvement in energy resolution could involve the same improvement in spatial resolution. In fact, the non-proportionality response of NaI(Tl) scintillation light with photon energy could affect the position response, as demonstrated from spatial resolution experimental variation of only 25% in the energy range of 70-511 keV. To this aim we realized three small gamma cameras based on LaBr3:Ce coupled to the Flat panel PMT. This detector optical configuration permits the narrowest light distribution and the highest light collection to obtain the best spatial and energy resolution values respectively.The crystal have the same detection areas (5 x 5 cm2) and various thickness (4 mm , 5 mm and 10 mm respectively). Ultimately, the detector responses have been compared with analogous ones obtained from a NaI(Tl) continuous crystal in the same detection assembly and with 1.5 mm thickness. All detectors have been scanned by gamma emitting radioactive sources collimated with 1 mm aperture diameter, in the energy range between 70 and 360 keV. Preliminary measurements confirm the linear dependence of spatial resolution on energy resolution providing better than 2 mm FWHM values at 245 keV photon energy with 10 mm crystal thickness. Energy resolution values were also measured by standard PMT, confirming the results previously reported for continuous LaBr3:Ce crystals. Excellent spatial resolution (around 1 mm) was obtained from LaBr3:Ce crystals of 4 and 5 mm thickness, at 140 keV photon energy, corresponding to a detection efficiency ranging between 70% and 80%. In conclusion, the results confirm that LaBr3:Ce gamma camera can give a consistent contribution to the imaging improvement in molecular field both for animal studies and for clinical applications.

Speaker: Prof. Roberto Pani (INFN Dept Experimental Medicine and Pathology - University of Rome "La Sapienza"-Italy)

14:45 New Developments in Compton SPECT and Compton PET Imaging

The Compton principle of electronic collimation overcomes the resolution-efficiency tradeoff imposed by mechanical collimators, and therefore Compton-based imagers may exhibit both high resolution and high efficiency. We have simulated and developed both Compton SPECT and Compton PET imaging prototypes. Simulations predict better than 3mm spatial resolution and about 10^-3 efficiency for a system consisting of an endorectal prostate probe with scintillator above and below a patient. A clinical prototype to validate the simulation results has been constructed. 1mm thick silicon pad sensors and associated readout electronics with energy resolution close to 1 keV FWHM have been developed. A resolution of 3.0mm FWHM at 3cm from the scatter detector was obtained with a Barium point source. A test bench was also developed to demonstrate the potential of solid state detectors for very high resolution small animal PET. A spatial resolution of 700 microns FWHM and a uniform response over a wide field of view was obtained. The time resolution of the solid-state detector as well as further improvements of the spatial resolution using a strong magnetic fields will also be presented.

Speaker: Prof. Harris Kagan (Dept. of Physics, Ohio State University)

15:00 Dedicated R&D for online PET imaging in hadrontherapy

It has been demonstrated [1] that PET information can be used successfully for online monitoring of particle range as well as for dose quantification in both proton and carbon therapy . Furthermore, it is clear that PET [1] is the only issue today for an in-situ and non-invasive treatment plan monitoring. However, existent PET systems are designed either for diagnostics or functional medical imaging where the emitters are injected in sufficient amounts to cope with the low sensitivity of the detectors. In hadrontherapy, the activity is due to rather short-lifetime target or projectile nuclear fragments like C, C or O. Thus, the activity is about two orders of magnitude lower than the corresponding activity in conventional PET imaging. The development of new efficient PET detectors dedicated to hadrontherapy becomes crucial. In this talk, after a brief description of PET requirements in hadrontherapy, we will describe our R&D activities dedicated to online plan control with PET in hadrontherapy. I will also present some preliminary results obtained with our prototype in real beam conditions. [1] K. Parodi, PhD thesis, FZR-415, Nov 2004, ISSN 1437-322X. See also the references therein.

Speaker: Prof. Madjid Boutemeur (Université Claude Bernard Lyon 1)

15:15 ENLIGHT – "The European Network for Light Ion Therapy"

Cancer is a major societal problem and it is the main cause of death between the ages of 45 to 65 years. In the treatment of cancer, radiotherapy (RT) plays an essential role. RT with ions, due to their unique physical and radiobiological properties, offers several advantages over photons. In particular, they penetrate the patient with minimal diffusion, they deposit their maximum energy at the end of their range, and they can be shaped as narrow focused and scanned pencil beams of variable penetration depth. Hadron beams allow highly conformal treatment (where the beam conforms to the shape of the tumour) of deep-seated tumours with great accuracy, while delivering minimal doses to surrounding tissues. RT with ions thus has great prospects for being used in early stages of tumour disease not amenable to surgery. It is likely that, besides its more impressive effect on radio- resistant tumours, post-treatment morbidity will be lower in patients treated with ions due to the lower dose and toxicity to normal tissues. Visionary physicist and founder of Fermilab Robert Wilson first proposed the use of hadrons for cancer treatment in 1946. This idea was first put into practise at the Lawrence Berkeley Laboratory (LBL) where 30 patients were treated with protons from 1954–1957. Since then the total number of patients treated with hadrons in the World now exceeds 48000 and almost 5000 new patients were treated last year. Several dedicated hospital-based centres with significant capacity for treating patients are now taking the place of the first R&D facilities hosted by the physics research laboratories (e.g. LBL, GSI). Europe is playing a key role in the development of light ion therapy facilities with five financed centres using actively scanned carbon ions (of which two are under construction in Heidelberg and Pavia) and several proton therapy centers which will become operational soon. In the US two proton therapy centers are running and four more are under construction. In the Far-East, in particular Japan but also Korea, and China are investing in hospital-based hadrontherapy centres. The European Network for Research in Light-Ion Hadron Therapy (ENLIGHT) was established in 2002 to co-ordinate European efforts in radiation therapy using light-ion beams. ENLIGHT has been instrumental in bringing together different European centres to promote hadron therapy, in particular with carbon ions. ENLIGHT created a multidisciplinary platform, uniting traditionally separate communities so that clinicians, physicists, biologists and engineers with experience in ions. The success of the network has encouraged the scientific community to promote more inclusive collaboration between the researchers and regional activities and to enlarge the collaboration to include the proton community. Hence ENLIGHT++ continues the vision started by ENLIGHT.

Speaker: Prof. Manjit Dosanjh (CERN)

15:30 → 16:00 Conclusion

15:30 Conclusion

Speaker: Dr William Moses