Presenter: Chris Layden
Authors: Chris Layden, Kyle Klein, William Matava, and Akhil Sadam
Research conducted through the University of Texas Karol Lang Laboratory (UTKL)
In the wake of advancements in scintillator and photodetector technologies, the medical imaging technique of positron emission tomography (PET) has seen considerable improvement in recent years. In particular, significant improvements have been made to the time resolution of PET scanners, allowing for tighter event localization and thus better signal-to-noise ratios. But optimization of time resolution conventionally requires short scintillator crystals, yielding diminished scanner sensitivity. A promising but clinically unrealized technique for striking a harmonious balance between sensitivity and time resolution is the coupling of a second photodetector to the front faces of long scintillator crystals. By comparing the signals from these two photodetectors for each event, one may predict the depth of gamma ray interaction in the scintillator crystal. One may then use this depth to correct the time of detection by subtracting the duration required for light to have traveled in the scintillator crystal. To study the benefits of double-ended readout, we first developed Monte Carlo simulations of a single-ended brain PET scanner consisting of Silicon Photomultiplier (SiPM) photodetectors and LYSO:Ce scintillator crystals of size 3x3x30 mm^3. These simulations were conducted using the Geant4 physics modeling toolkit. We found that this scanner would attain a coincidence resolving time (CRT) of 248±5 ps. We then developed simulations of the same scanner geometry with the addition of a SiPM to the front face of each scintillator crystal. With such double-ended readout, we found that the scanner would attain a much improved CRT of 171±5 ps. As this value far outpaces today's most advanced commercial PET scanners, the application of double-ended readout to clinical scanners would allow for a dramatic increase in patient throughput and a dramatic decrease in the administered dose required for each patient.
![](https://i.ytimg.com/vi/zRGFDtoF6SI/maxresdefault.jpg)