mDixon-Based Synthetic CT Generation for PET Attenuation Correction on Abdomen and Pelvis Jointly Using Transfer Fuzzy Clustering and Active Learning-Based Classification

• Published in: IEEE transactions on medical imaging Vol. 39; no. 4; pp. 819 - 832
• Main Authors: Qian, Pengjiang, Chen, Yangyang, Kuo, Jung-Wen, Zhang, Yu-Dong, Jiang, Yizhang, Zhao, Kaifa, Al Helo, Rose, Friel, Harry, Baydoun, Atallah, Zhou, Feifei, Heo, Jin Uk, Avril, Norbert, Herrmann, Karin, Ellis, Rodney, Traughber, Bryan, Jones, Robert S., Wang, Shitong, Su, Kuan-Hao, Muzic, Raymond F.
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.04.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Abdomen; Abdomen - diagnostic imaging; Active learning; active learning-based classification (ALC); Attenuation; attenuation correction (AC); Biological tissues; Bones; Classification; Clinical trials; Cluster Analysis; Clustering; Computed tomography; Dixon-based MR; Fuzzy Logic; Humans; Image classification; Image Processing, Computer-Assisted - methods; Learning algorithms; Machine learning; Magnetic Resonance Imaging; Medical imaging; Methods; Parameter robustness; Pelvis; Pelvis - diagnostic imaging; Positron-Emission Tomography - methods; Soft tissues; Support Vector Machine; Synthetic CT generation; Tissues; Tomography, X-Ray Computed; transfer fuzzy clustering (TFC); Transfer learning

We propose a new method for generating synthetic CT images from modified Dixon (mDixon) MR data. The synthetic CT is used for attenuation correction (AC) when reconstructing PET data on abdomen and pelvis. While MR does not intrinsically contain any information about photon attenuation, AC is needed...