Fifty-seventh annual meeting of the American association of physicists in medicine
TH-EF-BRB-04: One-Pixel Prototype of a Novel Polycrystalline CdTe-Based Multilayer Detector for Radiation Therapy Imaging Applications
Recent developments in polycrystalline CdTe-based photovoltaics have resulted in a technology of inexpensive, large-area structures possessing strong built-in fields and superior radiation hardness. Stacking a large number of such structures (number of layers N>>1) into a detector with signal readout from individual layers produces a device with sufficient absorption under an MV beam, low signal spreading, no charge trapping, and low noise due to obviating the need for external biasing. We tested one-pixel multilayer detector prototypes under a 6MV beam, compared their electrical parameters with those predicted by device modeling, and used the layer number-signal data for verification of the energy spectrum unfolding algorithm.
We modeled the energy deposition with depth of a CdTe-based device under a 6MV beam with the Monte Carlo simulation package MCNP5. The results yielded charge carrier generation profiles in individual layers, which were then combined into structures of 10–12 layers. The profiles were used as input for modeling of current-voltage (I-V) characteristics with the device operation modeling software SCAPS-1D. Modeling results were compared to measurements on several 1-pixel multilayer structures under the 6MV beam of a Varian TrueBeam accelerator.
Measurements of I-V characteristics for all layers of 1-pixel multilayer structures established electrical parameters of best devices to be close to those modeled. A regularization algorithm for extracting the energy spectrum was successfully tested with these structures.
The proposed multilayer structure adds a third dimension, depth-signal reading, to the traditional 2-D pixelation, allowing for the photon energy and source position to be obtained through signal processing. The latter processing relies on regularization approaches, similar to those developed for computed tomography, and thus have potential to significantly improve the quality of images under MV energy x-ray sources of medical linear accelerators.
This research is supported with NRC grant No. NRC-HQ-12-G-38-0042