No-reference evaluation of the reconstructed images in single-shot K-Edge Subtraction X-ray Computed Tomography

Single-shot K-Edge Subtraction X-ray Computed Tomography (CT) with a multi-threshold photon-counting detector is an interesting approach to favour low-dose analyses of a known contrast agent with promising applications in vivo. To assess the minimum detectable concentration of the contrast agent and to favour possible radiation dose reduction and/or faster acquisition time, a significant role is played by the tomographic reconstruction algorithm. By considering experimental images, this work evaluates three CT reconstruction methods and different acquisition statistics via a no-reference assessment of contrast-to-noise ratio and spatial resolution. The results support that, although computationally expensive, a SART-TV reconstruction approach yields adequate results even when a limited number of projections is available.