Transmission dosimetry via MAPS-based event discrimination in a medical linear accelerator

摘要

Transmission dosimetry, which places a detector upstream of the patient, is a promising method of dosimetry for intensity-modulated radiotherapy. Dose-predicted downstream of the patient is affected by attenuation and scattering. Changes in the radiation field are difficult to separate from changes in the beam that might indicate an erroneous treatment. Monolithic active pixel sensor (MAPS) technology offers a potentially cost-effective method for measuring the beam profile in transmission geometry. MAPS detectors can be made very thin (interacting with less than 1% of photons), while offering high spatial and temporal resolution, and therefore event shape and energy discrimination. This study examines event discrimination with MAPS detectors in the high dose-rate environment associated with a medical linear accelerator (LINAC), as a method of distinguishing between therapeutic and contaminant components of the radiation. Monte Carlo modeling of the LINAC and MAPS detector has been used to predict the spectrum of events in the panel, including an evaluation of the occupancy associated with individual LINAC pulses, occurring at rates of up to 40 Hz. It is shown via modeling and experimental measurements that there is sufficiently low occupancy to allow event discrimination. Identifiable signatures, including the dominance of photon events at low and high energy deposition limits, and features peculiar to contaminating charged-particles, suggest fruitful avenues. Loss of resolution, because of charge sharing in the detector observed in measurements of the LINAC beam, and increasingly higher dose-rates used in practice imply further challenges. Copyright (C) 2015 John Wiley & Sons, Ltd.