Cluster-Based Analytical Methods for Diffusion-Ordered NMR Spectroscopy and Apparent Diffusion Coefficient Mapping in MRI
Enping Lin,Liubin Wu,Yida Chen,Bo Chen,Jian Wu,Ze Fang,Haolin Zhan,Taishan Kang,Yuqing Huang,Yu Yang,Zhong Chen
DOI: https://doi.org/10.1109/tim.2024.3476616
IF: 5.6
2024-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:Diffusion is a vital molecular property exploited in the nuclear magnetic resonance (NMR) technique for component identification. Diffusion-ordered spectroscopy (DOSY) is a crucial diffusion-based analytical tool for identifying complex mixtures. Traditionally, DOSY relies on quantitative diffusion coefficient analysis, which normally requires tens of encoding gradients for precise component separation. However, under severe experimental conditions, where only a few gradients can be conducted, the data quality is too low for the existing methods to obtain robust quantitative estimations. To address this challenge, we introduce the cluster-based processing procedure named Cluster-DOSY. Unlike conventional methods, Cluster-DOSY employs a qualitative analysis technique, classifying molecular components based on the similarity of their decay signals. Experiments demonstrate that Cluster-DOSY not only facilitates easier DOSY implementation but also provides more robust separation results compared to existing quantitative methods, especially when limited encoding gradients are available. Finally, we also try to extend the proposed idea into a diffusion-based MRI application from the Cluster-DMRI procedure for tissue component separation analysis and discuss the advantages, limitations, and future improvements.