First application of a novel SRAM-based neutron detector for proton therapy

Abstract

In proton therapy, neutrons produced in collimators or in the patient body will contribute to unwanted additional radiation dose to the patient. This neutron dose is primarily associated with an increased risk of secondary cancer after treatment. Neutron detection in proton therapy has previously, to a large extent, been based on the use of passive detectors or detectors of large physical size which are not applicable for measurements inside phantoms. In this study, we present the first application of a neutron detector based on registration of radiation induced effects in Static Random Access Memories (SRAMs) which enables in-phantom fast neutron measurements. Measurements were performed in a water phantom irradiated by a 178 MeV proton pencil beam in a setup mimicking internally produced neutrons in pencil beam scanning (PBS) proton therapy. A neutron energy response model was developed for the SRAM detector to increase the accuracy of measurements in radiation fields with a broad range of neutron energies, and thereby make the detector applicable in the proton therapy setting. FLUKA Monte Carlo (MC) simulations and thermal neutron measurements with thermoluminescence detectors (TLDs) were conducted to evaluate the SRAM detector performance.Both experimental results (SRAM detector and TLDs) and MC simulations indicated a steep decrease in the neutron fluence with increasing lateral distance from the beam axis, while less variation was observed with depth. At Bragg peak depth, experimental values of fast neutron dose (H*(10)) were reduced from 1.4 mSv/Gy at 5.2 cm lateral distance to the beam axis to 0.22 mSv/Gy at 13.7 cm lateral distance. The MC simulations also showed that the SRAM detection threshold of 3 MeV was sufficiently low to capture approximately 90% of the neutron dose in PBS proton therapy. The differences in detector response due to variations in the neutron energy spectrum were small (<10%), and may be corrected using position-specific calibration factors. The results demonstrate the potential for estimating spatial out-of-field neutron dose based on the SRAM detector measurements in combination with the neutron energy response model.