An Iterative Dynamic Dual-Energy CT Model for Multi-Energy CT Imaging

Yidi Yao,Liang Li,Zhiqiang Chen

DOI: https://doi.org/10.1109/nss/mic44867.2021.9875604

2021-01-01

Abstract:With the progress of photon-counting detector, multi-energy CT imaging is attracting more attention in clinical diagnosis. However, the statistical noise increases with more energy bins, which degrades the image quality. Our team proposed the dynamic dual-energy CT scan mode for multi-energy imaging and multi-material decomposition, which enable multi-energy CT imaging with only two energy thresholds, and is effective in noise suppression. In this work, we propose a novel dynamic dual-energy CT model based on the dynamic dual-energy CT scan mode. In each iteration of the proposed CT model, the multi-energy projections are first estimated. Then the multi-energy CT images are reconstructed from the estimated projections. Compared to our previous work, the new proposed model introduced the estimation of multi-energy projections, simplifies and accelerates the reconstruction, and enables the possibility for further deep analysis of the dynamic dual-energy CT mode. The noise-free simulation result of the proposed model is extremely closed to the conventional multi-energy result, which illustrate the effectiveness of the work.