Iterative Clustering Material Decomposition Aided by Empirical Spectral Correction for High-Resolution Photon-Counting Detectors in Micro-CT (2310.10913v1)

Abstract: Photon counting detectors (PCDs) offer promising advancements in computed tomography (CT) imaging by enabling the quantification and 3D imaging of contrast agents and tissue types through multi-energy projections. However, the accuracy of these decomposition methods hinges on precise composite spectral attenuation values that one must reconstruct from spectral micro CT. Factors such as surface defects, local temperature, signal amplification, and impurity levels can cause variations in detector efficiency between pixels, leading to significant quantitative errors. In addition, some inaccuracies such as the charge-sharing effects in PCDs are amplified with a high Z sensor material and also with a smaller detector pixels that are preferred for micro CT. In this work, we propose a comprehensive approach that combines practical instrumentation and measurement strategies leading to the quantitation of multiple materials within an object in a spectral micro CT with a photon counting detector. Our Iterative Clustering Material Decomposition (ICMD) includes an empirical method for detector spectral response corrections, cluster analysis and multi-step iterative material decomposition. Utilizing a CdTe-1mm Medipix detector with a 55$\mu$m pitch, we demonstrate the quantitatively accurate decomposition of several materials in a phantom study, where the sample includes mixtures of material, soft material and K-edge materials. We also show an example of biological sample imaging and separating three distinct types of tissue in mouse: muscle, fat and bone. Our experimental results show that the combination of spectral correction and high-dimensional data clustering enhances decomposition accuracy and reduces noise in micro CT. This ICMD allows for quantitative separation of more than three materials including mixtures and also effectively separates multi-contrast agents.