An iterative-FBP dual-spectral CT reconstruction algorithm based on scatter modeling.

Abstract

Dual-spectral computed tomography (DSCT) has found extensive application in medical and industrial imaging due to its superior capability to distinguish different materials. However, a significant challenge in DSCT lies in the presence of X-ray scatter, which degrades the quality of reconstructed images. Traditional DSCT reconstruction methods often neglect the impact of scatter, leading to inaccurate basis material decomposition, especially under severe scatter conditions. To address this limitation, this paper proposes an innovative iterative DSCT reconstruction algorithm based on the filtered back-projection (FBP) method. Specifically, we first refine the commonly used polychromatic attenuation model to more accurately account for the effects of scatter. Building on this improved model, we propose an iterative reconstruction approach combined with the FBP method, achieving high-quality DSCT reconstructions that effectively mitigate scatter artifacts and improve the accuracy of basis material decomposition. Experiments on both simulated phantoms and real data demonstrate the superior performance of the proposed method in DSCT reconstruction. Notably, our approach achieves outstanding basis material decomposition results without requiring additional pre or post-processing steps, making it particularly suitable for practical DSCT applications.