Investigation of a Novel Snout-Mounted Quad-Compton Camera system for In Vivo Proton Dose Verification via Prompt Gamma Imaging

This study presents a novel simulation study for integrating four Compton Cameras (CC) into proton therapy systems by mounting them directly on the snout of the treatment gantry for prompt gamma imaging (PGI). This compact, fixed geometry simplifies detector integration, maintains alignment with the proton beam axis, and supports in vivo proton range verification. A Geant4-based Monte Carlo framework, PJ-MC, was used to simulate prompt gamma emissions (PGE) in water and patient-specific geometry. PG image reconstruction was performed using the Kernel Weighted Back Projection (KWBP) algorithm. To improve range accuracy, we introduced a post-processing method called Point Source Normalization (PSN), which uses precomputed PGI responses for monoenergetic gamma point sources at different depths to normalize PGI images and correct distortions from attenuation, geometric factors, and detector responses from PGE detected curve. Simulations with monoenergetic sources confirmed the influence of depth-dependent attenuation and inverse square losses, especially at higher proton energies. Clinical beam energies of 70, 150, and 200 MeV were tested. KWBP reconstructions showed increasing range discrepancies with energy, reaching 176 mm (water) and 144 mm (patient) at 200 MeV. PSN correction reduced these errors by up to 70%, bringing PG hotspots closer to the Bragg peak (Dmax). Simulations with anterior/posterior shifts showed that the PSN-corrected system could detect millimeter-scale range deviations from setup uncertainties and beam shifts. This is the first study to simulate and evaluate a snout-mounted CC system for PGI in proton therapy. Results support future validation under clinical beam conditions in phantoms and patients.