ROI-Wise Material Decomposition in Spectral Photon-Counting CT

TL;DR

This paper introduces a novel ROI-wise material decomposition method for spectral photon-counting CT that improves accuracy and reliability by optimizing basis materials through spatio-energy segmentation and sparsity regularization.

Contribution

The paper presents a new ROI-wise decomposition approach that enhances material differentiation in spectral CT by combining segmentation, optimized matrices, and sparsity constraints.

Findings

Higher accuracy compared to TV and L1-norm regularization methods.

Validated on both digital and physical data sets.

Significantly improved reliability in material decomposition.

Abstract

Spectral photon-counting X-ray CT (sCT) opens up new possibilities for the quantitative measurement of materials in an object, compared to conventional energy-integrating CT or dual energy CT. However, achieving reliable and accurate material decomposition in sCT is extremely challenging, due to similarity between different basis materials, strong quantum noise and photon-counting detector limitations. We propose a novel material decomposition method that works in a region-wise manner. The method consists in optimizing basis materials based on spatio-energy segmentation of regions-of-interests (ROIs) in sCT images and performing a fine material decomposition involving optimized decomposition matrix and sparsity regularization. The effectiveness of the proposed method was validated on both digital and physical data. The results showed that the proposed ROI-wise material decomposition method presents clearly higher reliability and accuracy compared to common decomposition methods based on total variation (TV) or L1-norm (lasso) regularization.