A simulation study of X-ray-induced acoustic dosimetry for a compact FLASH radiotherapy device using digital phantoms

FLASH radiotherapy (FLASH-RT) is a novel technique characterized by its capacity to deliver ultra-high radiation doses to tumor sites within fractions of a second, thereby effectively sparing normal tissues while maintaining therapeutic efficacy against tumors. The X-ray-based FLASH-RT holds particular promise for clinical adoption due to its accessibility to ultra-high-dose-rate beams, compact accelerator design, and reduced treatment costs. Accurate dosimetric verification during FLASH-RT delivery necessitates online monitoring and visualization of spatial dose distribution within the irradiation target. However, the extreme radiotherapy conditions pose substantial challenges to conventional dosimetry methods, such as signal saturation and measurement inaccuracies caused by pair production effects. To address these challenges, in this study, we propose and evaluate an X-ray-induced acoustic imaging (XAI) approach for in-phantom dosimetry in FLASH-RT. Using the Monte Carlo (MC) simulation toolkit Geant4, we modeled the irradiation of three heterogeneous tissue-mimicking phantoms with a single-pulse X-ray beam delivered by a 10 MV medical linear accelerator (Linac) integrated into a compact FLASH radiotherapy device. The generation, propagation, and attenuation of the induced acoustic waves following irradiation from the Linac were approximated using the k-space pseudospectral method. The acoustic signals were collected by a 128-element annular array transducer and subsequently used to reconstruct the spatial dose distribution within the phantoms. The results demonstrate the feasibility of XAI-based dosimetry for ultra-high-dose-rate X-ray beams even under low signal-to-noise ratio (SNR) conditions, providing a preliminary basis for its further development toward clinical FLASH-RT applications.