Speaker
Description
Positron emission tomography (PET) is the highest sensitivity technique available for imaging the entire human body. Because the axial field-of-view of state-of-the-art clinical scanners is ~25 cm, they surround only 2% solid angle of an object in the camera’s field of view (on average) and so are far from the full sensitivity potential. The goal of the EXPLORER project is to develop a total-body PET scanner for biomedical research whose 2 meter axial field-of-view pushes the limits of sensitivity and opens up a wealth of new possibilities for using PET to study health and disease. In addition to the geometric increase, another 2x effective sensitivity gain will be obtained by adding time-of-flight capability. The project is a multi-institutional collaboration funded by a NIH Transformative Research Award, and is intended to develop a unique and high performance research instrument for the entire nuclear medical imaging community.
The detector design is based on off-the-shelf, conventional LSO/LYSO- based block detectors used in high-performance commercial PET systems. Design goals include <4 mm spatial resolution, <400 ps timing resolution, <12% energy resolution, 76 cm ring diameter, and 200 cm axial field of view. GATE Monte Carlo simulations suggest noise-equivalent count improvements of ~40x for total- body imaging and 10x for single organ imaging, compared with current clinical PET scanners. The high efficiency along with the desire to image injected doses up to 30 mCi (1100 MBq) place challenging throughput requirements on the electronics. Thus, while real-time coincidence processing will be supported for lower injected doses, the system will also support singles mode data collection using a parallel architecture with coincidences formed offline in software.
We believe that the ~40x improvement in sensitivity of this camera, along with its ability to image the entire body simultaneously, will enable biomedical applications that are currently impossible. With the same PET protocols currently used (i.e., same radiotracer, injected dose, and imaging time), the signal to noise ratio in the reconstructed images should improve by a factor of >6. This improved SNR could be used to reconstruct images at higher resolution, enabling detection of smaller lesions / lower grade disease, and to reduce the statistical errors to enable more accurate kinetic modeling. Obtaining good quality images with 40x less activity would enable protocols that extend five additional half lives post-injection—3 additional hours for $^{11}$C imaging and 16 additional hours for $^{18}$F imaging. Much shorter scans also become possible—a full body scan that presently takes ~20 minutes (8 bed positions at 2–3 minutes per position) could be done in <30 seconds. Finally, the injected dose could be reduced 40x, yielding a radiation exposure similar to a round-trip trans-oceanic airline flight. This would enable longitudinal studies and possibly expand the use of PET into more radiation-sensitive populations.
