TY - GEN
T1 - Material decomposition in photon-counting-detector CT
T2 - Medical Imaging 2019: Physics of Medical Imaging
AU - Ren, Liqiang
AU - Tao, Shengzhen
AU - McCollough, Cynthia H.
AU - Yu, Lifeng
N1 - Funding Information:
Research reported in this publication was supported by the National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health under Award Numbers R21 EB024071, R01 EB016966, and C06 RR018898. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Publisher Copyright:
© SPIE. Downloading of the abstract is permitted for personal use only.
PY - 2019
Y1 - 2019
N2 - Energy-resolved photon-counting-detector CT (PCD-CT) is promising for material-specific imaging of multiple contrast agents. In each PCD-CT scan, two groups of images can be reconstructed, namely threshold images and bin images, and both can be directly used for material decomposition. The performance may differ for different energy thresholds and imaging tasks and it remains unclear which group of images should be used. The purpose of this work is to evaluate the imaging performance of threshold images and bin images when they are used for a three-material decomposition task (iodine, gadolinium, and water) in PCD-CT. Material decomposition was performed in image-space by using both an ordinary least squares (OLS) method and a generalized least squares (GLS) method. Both numerical analysis and phantom experiments were conducted, which demonstrated that: 1) compared with OLS, GLS provided improved noise properties using either threshold or bin images; 2) for the GLS method, when the covariances among images are taken into account, threshold and bin images showed almost identical material-specific imaging performance. This work suggested that, when correlations among images are incorporated into material decomposition, threshold and bin images perform equivalently well.
AB - Energy-resolved photon-counting-detector CT (PCD-CT) is promising for material-specific imaging of multiple contrast agents. In each PCD-CT scan, two groups of images can be reconstructed, namely threshold images and bin images, and both can be directly used for material decomposition. The performance may differ for different energy thresholds and imaging tasks and it remains unclear which group of images should be used. The purpose of this work is to evaluate the imaging performance of threshold images and bin images when they are used for a three-material decomposition task (iodine, gadolinium, and water) in PCD-CT. Material decomposition was performed in image-space by using both an ordinary least squares (OLS) method and a generalized least squares (GLS) method. Both numerical analysis and phantom experiments were conducted, which demonstrated that: 1) compared with OLS, GLS provided improved noise properties using either threshold or bin images; 2) for the GLS method, when the covariances among images are taken into account, threshold and bin images showed almost identical material-specific imaging performance. This work suggested that, when correlations among images are incorporated into material decomposition, threshold and bin images perform equivalently well.
KW - Generalized Least Squares (GLS)
KW - Material Decomposition
KW - Ordinary Least Squares (OLS)
KW - Photon-counting-detector CT (PCD-CT)
KW - Variance-covariance Matrix
UR - http://www.scopus.com/inward/record.url?scp=85068369377&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85068369377&partnerID=8YFLogxK
U2 - 10.1117/12.2513463
DO - 10.1117/12.2513463
M3 - Conference contribution
AN - SCOPUS:85068369377
T3 - Progress in Biomedical Optics and Imaging - Proceedings of SPIE
BT - Medical Imaging 2019
A2 - Schmidt, Taly Gilat
A2 - Chen, Guang-Hong
A2 - Bosmans, Hilde
PB - SPIE
Y2 - 17 February 2019 through 20 February 2019
ER -