Optimizing Transmittance for Enhanced Compressive X-Ray Compton Backscattering Imaging

The Compressive X-ray Compton Backscattering Imager (CXBI) is a method that overcomes the limitations of conventional pixel-by-pixel Compton scanning. It uses coded illumination projected onto the target, combined with relative movement between the mask and the body. Due to CXBI being recently proposed, limited efforts have been made to find optimal mask patterns and their respective parameters. In this paper, we present a data-driven framework to find optimal binary coding patterns for CXBI. This process uses compressive sensing regularizers that are directly related to the CXBI forward model. To find the optimal transmittance value (the percentage of non-blocking pixels in the pattern), the optimization problem was solved for various transmittance values. The results indicate that, on average, 10% and 30% transmittance yield the best Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Metric (SSIM) values.