The concept of structure engineering has been proposed for the exploration of the next generation of radiation detectors with improved performance. A Time Of Flight Positron Emission Tomography (TOF-PET) scanner with heterostructured scintillators with a pixel size of 3.0×3.1×15 mm3 was simulated. The heterostructures consisted of alternating layers of BGO as a dense material with high stopping power and plastic as a fast light emitter. Using the GATE simulation toolkit, a detector time resolution was calculated as a function of the deposited and shared energy in both materials on an event-by-event basis. We saw that while sensitivity was reduced to 32% for 100 µm thick plastic layers and 52% for 50 µm, the CTR distribution improved to 204±49 ps and 220±41 ps respectively, compared to 276±9 ps for bulk BGO. We divided the events into three groups based on their CTR and modeled them with different Gaussian TOF kernels. On a NEMA IQ phantom, the heterostructures had better contrast recovery in early iterations, while on the other hand, BGO achieved a better Contrast-to-Noise Ratio (CNR) after the 10th - 15th iteration due to the higher sensitivity. The developed simulation and reconstruction methods constitute new tools for evaluating different detector designs with complex time responses.

Today Time-Of-Flight in PET scanners assumes a unique, well-defined timing resolution for all coincidence events. However, recent BGO-Cherenkov detectors, combining prompt Cherenkov emission and the typical BGO scintillation signal, are capable of sorting events into multiple mixture timing kernels and therefore increase the available information for the image reconstruction. The number of Cherenkov photons detected per event impacts directly the signal rise time, which can therefore be used to improve the timing resolution. In this work, we present a simulation toolkit that outputs data with multiple timing resolutions and image reconstruction that incorporates this information. A full cylindrical BGO-Cherenkov PET model was compared, in terms of contrast recovery and contrast-to-noise ratio, against non-TOF and an LYSO model with time resolution of 213 ps. Two reconstruction approaches for the mixture kernels were tested; mixture Gaussian and decomposed simple Gaussian models. The decomposed model used the exact mixture component applied in the simulation. Images reconstructed using mixture kernels provided similar mean value and less noise as compared to the decomposed simple Gaussian kernels. Although, the later converged faster. Related to the standard LYSO model, the BGO-Cherenkov provided similar contrast, although, in most cases, with more iterations. However, due to the higher sensitivity, the contrast-to-noise ratio was 26.4% better for the BGO model.