Material Decomposition in Low-Energy Micro-CT Using a Dual-Threshold Photon Counting X-Ray Detector

Research output: Contribution to journalArticlepeer-review

Abstract

Material decomposition in computed tomography is a method for differentiation and quantification of materials in a sample and it utilizes the energy dependence of the linear attenuation coefficient. In this study, a post-image reconstruction material decomposition method is constructed for a low-energy micro-CT setup using a photon counting x-ray detector. The low photon energy range (4–11 keV) allows for K-edge contrast separation of naturally occurring materials in organic tissue without the need of additional contrast agents. The decomposition method was verified using a phantom and its capability to decompose biomedical samples was evaluated with paraffin embedded human atherosclerotic plaques. Commonly, the necessary dual energy data for material decomposition is obtained by manipulating the emitted x-ray spectrum from the source. With the photon counting detector, this data was obtained by acquiring two energy window images on each side of the K-edge of one material in the sample. The samples were decomposed into three materials based on attenuation values in manually selected regions. The method shows a successful decomposition of the verification phantom and a distinct distribution of iron, calcium and paraffin in the atherosclerotic plaque samples. Though the decompositions are affected by beam hardening and ring artifacts, the method shows potential for spectral evaluation of biomedical samples.
Original languageEnglish
Article number673843
Number of pages8
JournalFrontiers in Physics
Volume9
DOIs
Publication statusPublished - 2021 May 10

Subject classification (UKÄ)

  • Accelerator Physics and Instrumentation
  • Atom and Molecular Physics and Optics
  • Other Physics Topics

Fingerprint

Dive into the research topics of 'Material Decomposition in Low-Energy Micro-CT Using a Dual-Threshold Photon Counting X-Ray Detector'. Together they form a unique fingerprint.

Cite this