Abstract
CONTEXT: Using prompt gamma (PG) ray is proposed as a promising solution for in vivo monitoring in proton therapy. Despite significant and diverse approaches explored over the past two decades, challenges still persist for more effective utilization. AIMS: The feasibility of estimating proton range with PG imaging (PGI) as an online imaging guide in an anthropomorphic phantom with lung cancer was investigated through GATE/GEANT4 Monte Carlo simulation. SETTING AND DESIGN: Once the GATE code was validated for use as a simulation tool, the gamma energy spectra of NURBS-based cardiac-torso (NCAT) and polymethyl methacrylate phantoms, representing heterogeneous and homogeneous phantoms respectively, were compared with the gamma emission lines known in nuclear interactions with tissue elements. A 5-mm radius spherical tumor in the lung region of an NCAT phantom, without any physiological or morphological changes, was simulated. SUBJECTS AND METHODS: The proton pencil beam source was defined as a function of the tumor size to encompass the tumor volume. The longitudinal spatial correlation between the proton dose deposition and the distribution of detected PG rays by the multi-slit camera was assessed for proton range estimation. The simulations were conducted for both 10(8) and 10(9) protons. RESULTS: The deviation between the proton range and the range estimated by PGI following proton beam irradiation to the center of the lung tumor was determined by evaluating the longitudinal profiles at the 80% fall-off point, measuring 1.9 mm for 10(9) protons and 4.5 mm for 10(8) protons. CONCLUSIONS: The accuracy of proton range estimation through PGI is greatly influenced by the number of incident protons and tissue characteristics. With 10(9) protons, it is feasible to utilize PGI as a real-time monitoring technique during proton therapy for lung cancer.