Renkuan Zhai1, Xiaoqian Huang2, Yawei Zhao1, Meiling Ji1, Xuyang Lv2, Mengyao Qian1, Shu Liao2, and Guobin Li1
1United Imaging Healthcare, Shanghai, China, 2United Imaging Intelligence, Shanghai, China
1United Imaging Healthcare, Shanghai, China, 2United Imaging Intelligence, Shanghai, China
AI-assisted Compressed Sensing is introduced to realize superior MR acceleration and address the uncertainty of CNN by incorporating the output
of AI as one more regularization term.
Figure 1. ACS flowchart: AI Module represents
the trained AI network; CS Module is for solving the second formula using iterative
reconstruction.
Figure 2. T2 FSE 2D of knee with acceleration factors from 2.00x to 4.00x.
As acceleration factor increased, quality of CS images degraded more rapidly than
ACS.