CIF:Small:Coded Aperture Spectral X-Ray Tomography

University Of Delaware, Newark DE

Investigators

Abstract

Spectral tomography has emerged as a key imaging modality for medical imaging and homeland security. It offers material discrimination providing useful cues for identifying atomic composition, chemical targets, and explosives. It relies on the principle that the X-ray absorption of different chemical materials has characteristics that depend on the energy of the X-rays used. Thus, detectors that can differentiate between X-ray energies make it possible to discriminate different materials having similar X-ray attenuation properties. Unfortunately, energy resolving X-ray detectors are prohibitively costly for many applications. Other detector technologies often trade cost for acquisition speed and, in many cases, practical limitations require limited angle or subsampled projection geometries that lead to severe artifacts in image reconstruction. This research explores coded X-ray illumination projections which can be captured with standard integrating detectors yet provide information for energy-resolved image reconstructions. The coded measurements have lower radiation dose. They can also be multiplexed for robustness in limited angle scenarios. This research explores colored X-ray coded projections to circumvent current detector limitations. Rather than measuring objects by the attenuated X-rays, the X-rays are first coded in amplitude and spectra that modulate the X-ray field in a known pattern, effectively creating lower-dose structured X-ray bundles that interrogate specific voxels of the object. Careful design of the coded apertures, their placement, and the development of fast computational inverse algorithms that exploit the structured illumination can provide advantages in the application of spectral tomography. Various computerized tomography (CT) geometries are being considered including tomosynthesis, fan-beam, and cone-beam. A test-bed for validating the underlying theoretical concepts is a central part of the project.