Wenbo Liu1,2, Niranjan Balu2,
Xihai Zhao1, Huijun Chen2, Chun Yuan1,2,
Huilin Zhao3, Jianrong Xu3, Guangzhi Wang1,
William S. Kerwin2
1Biomedical Engineering
& Center for Biomedical Imaging Research, School of Medicine, Tsinghua
University, Beijing, China, People's Republic of; 2Radiology,
University of Washington, Seattle, WA, United States; 3Radiology,
Shanghai Jiao Tong University, Shanghai, China, People's Republic of
To evaluate the use of 3D gradient echo imaging techniques to identify plaque components, we sought to compare performance of an automatic segmentation algorithm based on Morphology-Enhanced Probabilistic Plaque Segmentation (MEPPS) using traditional 2D spin echo images versus 3D gradient echo images. The results show in principle a protocol based on 3D gradient echo contrast weightings provides sufficient information to quantitatively characterize calcification, necrotic core, and loose matrix in carotid atherosclerotic plaque. The advantage of this approach is especially well represented by the 3D visualization results. The isotropic segmentation results permit visualization of plaque structure in arbitrarily reformatted cuts.