Hongyu Li1, Mingrui Yang2, Jeehun Kim2, Chaoyi Zhang1, Ruiying Liu1, Peizhou Huang3, Sunil Kumar Gaire1, Dong Liang4, Xiaoliang Zhang3, Xiaojuan Li2, and Leslie Ying1,3
1Electrical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 2Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 3Biomedical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 4Paul C. Lauterbur Research Center for Biomedical Imaging, Medical AI research center, SIAT, CAS, Shenzhen, China
1Electrical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 2Program of Advanced Musculoskeletal Imaging (PAMI), Cleveland Clinic, Cleveland, OH, United States, 3Biomedical Engineering, University at Buffalo, State University of New York, Buffalo, NY, United States, 4Paul C. Lauterbur Research Center for Biomedical Imaging, Medical AI research center, SIAT, CAS, Shenzhen, China
This
abstract presents a combined deep learning framework to generate MR parameter
maps from very few subsampled echo images.
FIGURE 1. Schematic comparison of the
conventional model fitting and combined deep learning network SuperMAP with
joint spatial-temporal under-sampling.
FIGURE 2. T1rho maps from 3 echoes using combined
network SuperMAP and with single CNN network (RF 10.66), and the reference
T1rho maps from eight fully sampled echoes.