Long‐term beam output stability of an accelerator‐based boron neutron capture therapy system

Background Accelerator‐based boron neutron capture therapy (AB‐BNCT) systems are becoming commercially available and are expected to be widely used in hospitals. To ensure the safety of BNCT, establishing a quality assurance (QA) program and properly managing the stability of the system are necessary. In particular, a high level of beam output stability is required to avoid accidents because beam output is a major factor in patient dose. However, no studies have analyzed the long‐term beam output stability of AB‐BNCT systems. Purpose This study aimed to retrospectively analyze the long‐term stability of the beam output by statistical process control (SPC) based on the QA results over 3 years. Methods The data analyzed are the results of daily QA (DQA) and weekly QA (WQA) in an AB‐BNCT system and were taken between June 2020 and September 2023. The evaluation of the stability of the beam output was based on the reaction rate between gold and neutrons calculated using the activation foil method using a gold foil. In DQA, which can be performed in a short time, the gold foil was applied directly to the beam irradiation aperture in air. In WQA, measurements were performed at the phantom surface, 2‐cm depth, and 6‐cm depth using a dedicated water phantom. The acquired data were retrospectively analyzed by individuals and a moving range chart (I‐MR chart), exponentially weighted moving average control chart (EWMA chart), and several process capability indexes (PCIs). Results Over 99% of the DQA I‐MR chart results were within control limits, whereas the WQA I‐MR chart results showed that 1.8%, 4.1%, and 2.0% of the measurements exceeded the control limits at the surface, 2‐cm depth, and 6‐cm depth, respectively. The variation in the reaction rate of the gold foil before and after the replacement of the target was 1.0; however, the Cpmk of DQA and the 2‐ and 6‐cm depth WQAs between August 2021 and November 2022 in treatment course 2 were 0.83, 0.77, and 0.87, respectively, which were