Feng Cheng1, Yong Chen2, and Pew-Thian Yap3
1Department of Computer Science, University of North Carolina, Chapel Hill, NC, United States, 2Case Western Reserve University, Cleveland, OH, United States, 3Department of Radiology and Biomedical Research Imaging Center, University of North Carolina, Chapel Hill, NC, United States
1Department of Computer Science, University of North Carolina, Chapel Hill, NC, United States, 2Case Western Reserve University, Cleveland, OH, United States, 3Department of Radiology and Biomedical Research Imaging Center, University of North Carolina, Chapel Hill, NC, United States
We developed a deep learning method for rapid high-resolution 3D MRF with 16x acceleration. Whole-brain 3D MRF with 0.8 mm isotropic resolution can be achieved within 5 min acquisition time, making simultaneous T1 and T2 quantification possible in clinical settings.
Figure 4 Representative T1 and T2 maps obtained using retrospective 4x acceleration in k-space and 4x in image space. The proposed method achieves better performance both quantitatively and qualitatively.
Figure 1 Method overview. Our method consists of a Graph Convolutional Network for k-space and a Quantification Network for image space for reconstruction from undersampled 3D MRF data.