Library
home (Home) TU Delft Repository login Login

Accelerated white matter lesion analysis based on simultaneous T1 and T2∗ quantification using magnetic resonance fingerprinting and deep learning

Journal Article (2021)
Author(s)

Ingo Hermann (TU Delft - ImPhys/Computational Imaging, University Heidelberg)

Eloy Martínez-Heras (Universitat Autònoma de Barcelona)

Benedikt Rieger (University Heidelberg)

Ralf Schmidt (University Heidelberg)

Alena Kathrin Golla (University Heidelberg)

Jia Sheng Hong (National Taipei University of Technology)

Wei Kai Lee (National Taipei University of Technology)

Martijn Nagtegaal (TU Delft - ImPhys/Computational Imaging, TU Delft - ImPhys/Medical Imaging)

Sebastian Weingärtner (TU Delft - ImPhys/Computational Imaging, TU Delft - ImPhys/Medical Imaging)

undefined More Authors (External organisation)

DOI related publication
https://doi.org/10.1002/mrm.28688 Final published version
More Info
expand_more
Publication Year
2021
Language
English
Issue number
1
Volume number
86
Pages (from-to)
471-486
Downloads counter
362
Collections
Institutional Repository
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Purpose: To develop an accelerated postprocessing pipeline for reproducible and efficient assessment of white matter lesions using quantitative magnetic resonance fingerprinting (MRF) and deep learning. Methods: MRF using echo-planar imaging (EPI) scans with varying repetition and echo times were acquired for whole brain quantification of (Formula presented.) and (Formula presented.) in 50 subjects with multiple sclerosis (MS) and 10 healthy volunteers along 2 centers. MRF (Formula presented.) and (Formula presented.) parametric maps were distortion corrected and denoised. A CNN was trained to reconstruct the (Formula presented.) and (Formula presented.) parametric maps, and the WM and GM probability maps. Results: Deep learning-based postprocessing reduced reconstruction and image processing times from hours to a few seconds while maintaining high accuracy, reliability, and precision. Mean absolute error performed the best for (Formula presented.) (deviations 5.6%) and the logarithmic hyperbolic cosinus loss the best for (Formula presented.) (deviations 6.0%). Conclusions: MRF is a fast and robust tool for quantitative (Formula presented.) and (Formula presented.) mapping. Its long reconstruction and several postprocessing steps can be facilitated and accelerated using deep learning.