Euromedim 2006 :1st European Conference on Molecular Imaging Technology

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

9 May 2006, 14:00

1h

Palais du Pharo, Marseille

poster • Modeling, sampling strategies Poster Session :Simulation, Modeling, Reconstruction

Speaker

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

Description

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.