Characterization of the signal response of a prototype amorphous silicon active matrix flat panel dosimeter

The data reduction algorithms used for active matrix flat panel imagers (AMFPIs) require that a dark frame taken prior to x-ray exposure be subtracted from a radiation frame to remove offset and dark current components. This reduction model is linear and assumes that the dark current contribution to a radiation frame is equal to the dark current obtained from a dark data frame. This work investigates the behavior of a prototype active matrix flat panel dosimeter (AMFPD), based on an AMFPI, operated radiographically in a direct detection mode without a scintillator for dosimetric measurements. The experiments described were performed with the AMFPD at 10 cm deep in solid water with a source to surface distance of 90 cm. Measurements were made at dose rates ranging from 80 cGy/min to 320 cGy/min with total doses varied from 0.4 cGy to 300 cGy for a 6 MV beam from a linear accelerator. Field sizes ranged from 5 times 5 cm² to 20 times 20 cm². For long frame times and high exposures, a linear data reduction model (similar to the model used for image reconstruction in AMFPIs) is incorrect due to the loss of signal incurred in the radiation frame. Dose and dark current signals are decreased because of charge accumulation within the volume of the detector material (amorphous silicon) due to charge trapping. This effect can exceed 20%, which is dosimetrically unacceptable. To account for the nonlinear behavior, a reconstruction algorithm was developed with one free parameter. The free parameter is the effective dielectric constant for the detector material. A model based on the loss of the effective bias voltage across the detector is proposed. Validation of this model was performed by a direct comparison of measured data from the detector under this model with data from ion chamber scanned dose profiles. Finally, the consequences of this non-linear model for the use of an AMFPD operating in radiographic mode are discussed.