Abstract
Boron neutron capture therapy (BNCT) is a highly targeted radiation therapy that shows great promise for treating tumors that are challenging to address with conventional methods. The dose deposited in the tumor during a treatment can be monitored by detecting prompt gamma rays at 478 keV generated by boron neutron capture reactions within the tumor cells. However, this task is highly challenging due to the significant background of neutrons and gamma rays present during treatment that risk to mask the useful signal. An additional challenge is represented by borated polyethylene typically used for radioprotection purposes in the walls of the treatment rooms, which generates gamma rays of the same energy of the ones of clinical interest. To address these issues, we propose a scintillator-based detection system, integrating a pinhole collimator, an artificial neural network for gamma-ray position reconstruction and a multi-layer shielding strategy. This system successfully imaged borated samples with concentrations as low as 1843 ppm of (10)B, achieving a spatial resolution of approximately 1 cm, during neutron irradiation with a fluence rate of 10(7) n/cm(2)/s at the accelerator-based neutron facility at Nagoya University, demonstrating its potential for dose monitoring in clinical-like BNCT environments.