基于深度学习的快速磁共振CEST定量方法及成像设备
By using deep learning to predict the reference z-spectrum and optimize frequency offset selection, the problems of long imaging time and slow calculation speed in CEST imaging technology are solved, achieving rapid and accurate CEST quantification, which has good clinical application value.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2025-10-31
- Publication Date
- 2026-07-17
AI Technical Summary
Existing CEST imaging technology suffers from long imaging time and slow calculation speed during quantitative signal acquisition and post-processing. In particular, the computational intensity and long processing time of the NEMR method limit its clinical application.
We employ a fast magnetic resonance CEST quantitative method based on deep learning. By predicting the reference z-spectrum through a deep neural network group and combining it with a BM equation decoder, we optimize the frequency shift selection, reduce the number of necessary frequency shifts, and improve computational speed and imaging efficiency.
It achieves rapid and accurate CEST quantification, significantly improves calculation speed and imaging time, has good clinical application value, and reduces imaging time and post-processing calculation time.
Smart Images
Figure CN121385757B_ABST