<b>Objective:</b> To demonstrate the potential of PET imaging using scattered photons, we have proposed a novel technique that utilizes single-tissue-scattered events based solely on time-of-flight (TOF) information, without relying on energy data. Additionally, we explored the possibility of improving image quality by applying polarization selection criteria for true events.<br><b>Methods:</b> Due to Compton scattering within the phantom, scattered photons exhibit reduced energy, which remains unknown in our analysis. For a fixed scattering angle, the locus of the scattering point forms an arc of a circle in two dimensions (2D). With known TOF, we can trace the locus of the annihilation point for all possible scattering angles and positions. Our procedure involves identifying the annihilation-point loci corresponding to all pairs of single-scattered photons and merging them. In a separate study, we utilized the classical concept of orthogonal polarization of two annihilation photons, combined with energy filtering, to select genuine (true) events.<br><b>Results:</b> Our proof-of-concept study successfully demonstrated that the TOF-based approach could yield an image of a point source. The merging of annihilationpoint loci from numerous pairs of single-scattered photons produced a localized region for the source activity. The intensity profile showed a finite width of 14 mm for 1 ps uncertainty in TOF. Additionally, another analysis utilizing polarization and energy criteria generated a clean dataset for true events, with minimal background interference (containing random coincidences) at a high event rate.<br><b>Conclusions:</b> The simulation-based study validated our proposed model, indicating its potential to enhance the sensitivity and accuracy of event selection in PET imaging. These findings lay a solid foundation for further research into advanced PET imaging techniques that incorporate single-scattered events.
Read full abstract