Leveraging ultra-fast Cherenkov emission in scintillator-based TOF-PET by exploiting photon wavelength classification

Project Summary (Abstract) We propose to separate scintillation and Cherenkov photons produced in scintillator crystals to improve the time and energy resolution of time-of-flight positron emission tomography (TOF-PET) detectors far beyond those achieved in state-of-the-art systems. With its pico-molar sensitivity and a few millimeters of spatial resolution,

TOF-PET is the leading nuclear imaging modality for a number of diseases, from cancer to neurological and cardiovascular disorders. A significant improvement of the coincidence time resolution (CTR) and energy resolution would boost the signal-to-noise ratio and hence enhance image quality, resulting in more accurate

diagnoses, lower patient doses and exposure times, and granting access to a new broad range of applications for TOF-PET. The ultra-fast picosecond emission of Cherenkov light has demonstrated to achieve the best CTR ever reached of 30ps FWHM using PbF2, a pure Cherenkov emitter. However, this provides a very poor energy

resolution due to the low light yield of Cherenkov emission. The combination of Cherenkov and scintillation emission has been proposed as a way to obtain both good time and energy resolution, which has been demonstrated in bismuth germanium oxide (BGO), a high stopping power scintillator for PET, to obtain a CTR of

120ps FWHM with an energy resolution of 14%. The main reason why it is very challenging for BGO to reach CTRs of 30ps FHWM is due to the presence of the slower scintillation light and the inability of current detectors to disentangle between Cherenkov and scintillation. Additionally, the difference between the Cherenkov and

scintillation light emission spectra, makes it very hard to obtain a BGO detector that provides both good time and energy resolution. We propose to separate Cherenkov and scintillation photons in order to provide a detector that can be optimized independently for each of the signals, maximizing time resolution with Cherenkov and

energy resolution with scintillation without hindering each other. This separation can be achieved by exploiting the different emission spectra of each mechanism using dichroic filters, which are able to classify photons by wavelength with a negligible photon loss. This project aims to 1) obtain a CTR of 50ps FWHM and reduce the

scintillation background by a factor of 5 through wavelength classification in BGO, 2) increase photon detection efficiency in BGO by at least a factor of 2 without compromising time resolution, and 3) reach a CTR of 30ps FWHM with a 7% energy resolution by leveraging the hybrid Cherenkov-scintillation

concept with thallium chloride (TlCl). This project will pioneer the exploration of wavelength information as a way to dramatically improve TOF-PET performance. We will combine this technique with other cutting-edge technologies such as fast or high quantum efficiency photosensors, in order to demonstrate a novel a cost-

effective approach to a next generation TOF-PET. Our goal is to enable a new technology that can bring CTR closer to the 10ps FWHM milestone with a good energy resolution in order to be further exploited in future projects for the construction of a full TOF-PET system.