Advances in 3D Source Reconstruction Using Coded-Aperture Imaging

Gamma-ray imagers are useful for identifying the angular location of radioactive sources. Although the typical angular resolution of a coded-aperture gamma-ray imager is quite good (approximately 5 degrees), depth information is uncertain in the near field and essentially unknown in the far field. Many applications are aimed at quantifying the amount of source material being imaged, and to that end, knowledge of the depth of a source is required. Depth information, and thus the full 3D location, of a source can be acquired by incorporating measurements from imagers with multiple lines of sight. Iterative methods, such as Maximum Likelihood, provide a framework for 3D source reconstruction using multiple imager views. This paper discusses recent advances in applying a Maximum Likelihood iterative method to reconstruct radioactive sources in 3D using coded-aperture gamma-ray imagers. Advances include data-driven convergence criteria, extended source reconstruction in 2D, background subtraction assessments for various signal-to-noise-ratios, and point source reconstruction in 3D.