Prompt gamma ray detection and imaging for boron neutron capture therapy using CdTe detector and novel detector shield – Monte Carlo study

Background: For boron neutron capture therapy (BNCT), the improvements in patient dosimetry will require information about the spatial variation of ¹⁰B concentration in the tumor and critical organs. A non-invasive approach, based on the detection of prompt gamma (PG) rays from the BNC reaction, may be well-suited to obtain such information. The detectability of the BNC PG rays has been shown experimentally utilizing energy-resolving cadmium telluride (CdTe) detectors. However, the feasibility of this approach under the clinically relevant conditions of BNCT is currently unknown. Purpose: The present work aimed to investigate the aforementioned feasibility by performing Monte Carlo (MC) simulations under the phantom irradiation geometry relevant to accelerator-based BNCT (a-BNCT). Especially, this investigation focused on demonstrating the enhanced detection of the BNC PG rays using a novel neutron shield for CdTe detectors. Upon demonstrating the efficacy of the proposed detector shield, the BNC PG ray-based quantitative imaging of clinically relevant concentrations of ¹⁰B was also demonstrated. Methods: The Geant4 MC simulation toolkit was used to model the phantom irradiation by an epithermal neutron beam as well as the detection of the BNC PG rays from the phantom by CdTe detectors with and without the proposed gadolinium (Gd)-based detector shield. It was also used to model the BNC PG ray-based quantitative imaging of ¹⁰B concentrations under a-BNCT scenarios. Each model included a 20 cm-diameter/24 cm-height cylindrical PMMA phantom containing ¹⁰B inserts at various concentrations. Arrays of CdTe crystals of 5 × 5 × 1 mm³ each (up to 120 in the case of a ring detector) were modeled for acquiring the BNC PG ray signals and quantitative imaging. Results: According to the MC simulations, thermalized neutrons from the phantom were found to reach the CdTe detector and captured by Cd and Te, resulting in the gamma ray background noise that directly interfered with the BNC PG ray signal. The proposed Gd-based detector shield was found to be highly effective in shielding thermal neutrons from the phantom, thereby reducing the unwanted gamma ray background noise. Owing to this shield, the detection of as low as seven parts-per-million (ppm) of ¹⁰B within the phantom of clinically relevant size was possible using 20 billion incident neutron histories. Furthermore, quantitative imaging of ¹⁰B distributed at low concentration (down to 50 ppm) within the phantom was demonstrated using computed tomography (CT) simulations with 16 billion incident neutron histories per angular projection. The ¹⁰B detection limit (7.5 ppm) was also estimated using the reconstructed CT image. Both ¹⁰B detection limits determined from this investigation are deemed clinically relevant for BNCT. Conclusions: The proposed Gd-based detector shield played an essential role for achieving the currently reported ¹⁰B detection limits. Overall, the present MC simulation work demonstrated highly sensitive BNC PG ray detection and imaging under a-BNCT scenarios using CdTe detectors coupled with a novel detector shield.