Timing resolution is one of the most important features of current and future radiation detectors for diagnostic and therapeutic imaging. In positron emission tomography (PET), it has drastically improved image quality through time-of-flight (TOF) with a resolution of 350-500 ps that allows for the localization of positron annihilations? a direct measure of the activity distribution in the patient, with an uncertainty of 5-7.5 cm. Further improvement in image contrast could be obtained, with the ultimate goal of directly reconstructing the positron annihilation through an ambitious target of 10 ps timing resolution. Improving the detector timing requires the light transport to be thoroughly optimized, which can only be done through accurate Monte Carlo simulation. This proposal will develop accurate optical simulation tools for nuclear medicine detectors and will apply them to the design of fast detectors for TOF PET. The opensource software GATE and Geant4 constitute the main simulation platform in nuclear imaging and therapy. It includes optical transport in scintillators, but the models used to describe the light reflecting on the scintillator surfaces are highly inaccurate. We have developed and integrated into GATE a new optical model, the ?LUT Davis model? that addresses this limitation. This work, supported by an NIH R03 grant, demonstrated the feasibility of accurate scintillator optical modeling and opened the possibilities of using simulations for detector timing studies. Of particular interest are Cerenkov photons that are being investigated to improve timing resolution as low as 10 ps, which will require more advanced simulation tools. We will first develop and freely distribute computational tools to generate custom optical surface LUTs This aim is expected to have a strong impact on the nuclear imaging community where new detectors are being designed for future generations of scanners. Second, we will specifically develop optical models for photon timing studies, with the goal of establishing a comprehensive simulation framework for detector timing optimization. Third, we will apply our advanced Monte Carlo simulation tools to optimize the use of Cerenkov photons for a cost-effective BGO TOF PET detector and to develop the first semiconductor TOF PET detector. With these timing studies, we will tackle one of the most important challenges in PET research, which has the potential to transform PET instrumentation.
Nuclear medicine is a powerful tool for diagnostic and research in cancer, neurodegenerative and cardiovascular diseases. Current research on improving nuclear imaging instrumentation focuses on developing fast detectors such as time-of-flight detectors for positron emission tomography and relies on accurate simulation studies. The goal of this work is to implement new optical simulation tools for nuclear medicine and apply them to develop the next generation of fast and ultra-fast detectors.