Using deep learning to accelerate magnetic resonance measurements of molecular exchange

Zhaowei Cheng; Songtao Hu; Guangxu Han; Ke Fang; Xinyu Jin; Alfredo Ordinola; Evren Özarslan; Ruiliang Bai

doi:10.1063/5.0159343

Using deep learning to accelerate magnetic resonance measurements of molecular exchange

J Chem Phys. 2023 Aug 7;159(5):054201. doi: 10.1063/5.0159343.

Authors

Zhaowei Cheng¹, Songtao Hu², Guangxu Han², Ke Fang¹, Xinyu Jin¹, Alfredo Ordinola³, Evren Özarslan³, Ruiliang Bai^{2

4

5}

Affiliations

¹ College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou, China.
² Key Laboratory of Biomedical Engineering of Education Ministry, College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, China.
³ Department of Biomedical Engineering, Linköping University, Linköping, Sweden.
⁴ Interdisciplinary Institute of Neuroscience and Technology, School of Medicine, Zhejiang University, Hangzhou, China.
⁵ Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-Machine Integration, State Key Laboratory of Brain-Machine Intelligence, Zhejiang University, 1369 West Wenyi Road, Hangzhou 311121, China.

PMID: 37548304
DOI: 10.1063/5.0159343

Abstract

Real-time monitoring and quantitative measurement of molecular exchange between different microdomains are useful to characterize the local dynamics in porous media and biomedical applications of magnetic resonance. Diffusion exchange spectroscopy (DEXSY) is a noninvasive technique for such measurements. However, its application is largely limited by the involved long acquisition time and complex parameter estimation. In this study, we introduce a physics-guided deep neural network that accelerates DEXSY acquisition in a data-driven manner. The proposed method combines sampling pattern optimization and physical parameter estimation into a unified framework. Comprehensive simulations and experiments based on a two-site exchange system are conducted to demonstrate this new sampling optimization method in terms of accuracy, repeatability, and efficiency. This general framework can be adapted for other molecular exchange magnetic resonance measurements.