Literature DB >> 32767489

Deep learning using a biophysical model for robust and accelerated reconstruction of quantitative, artifact-free and denoised R 2 * images.

Max Torop1, Satya V V N Kothapalli2, Yu Sun1, Jiaming Liu3, Sayan Kahali2, Dmitriy A Yablonskiy2, Ulugbek S Kamilov1,3.   

Abstract

PURPOSE: To introduce a novel deep learning method for Robust and Accelerated Reconstruction (RoAR) of quantitative and B0-inhomogeneity-corrected R 2 * maps from multi-gradient recalled echo (mGRE) MRI data.
METHODS: RoAR trains a convolutional neural network (CNN) to generate quantitative R 2 ∗ maps free from field inhomogeneity artifacts by adopting a self-supervised learning strategy given (a) mGRE magnitude images, (b) the biophysical model describing mGRE signal decay, and (c) preliminary-evaluated F-function accounting for contribution of macroscopic B0 field inhomogeneities. Importantly, no ground-truth R 2 * images are required and F-function is only needed during RoAR training but not application.
RESULTS: We show that RoAR preserves all features of R 2 * maps while offering significant improvements over existing methods in computation speed (seconds vs. hours) and reduced sensitivity to noise. Even for data with SNR = 5 RoAR produced R 2 * maps with accuracy of 22% while voxel-wise analysis accuracy was 47%. For SNR = 10 the RoAR accuracy increased to 17% vs. 24% for direct voxel-wise analysis.
CONCLUSIONS: RoAR is trained to recognize the macroscopic magnetic field inhomogeneities directly from the input magnitude-only mGRE data and eliminate their effect on R 2 ∗ measurements. RoAR training is based on the biophysical model and does not require ground-truth R 2 * maps. Since RoAR utilizes signal information not just from individual voxels but also accounts for spatial patterns of the signals in the images, it reduces the sensitivity of R 2 * maps to the noise in the data. These features plus high computational speed provide significant benefits for the potential usage of RoAR in clinical settings.
© 2020 International Society for Magnetic Resonance in Medicine.

Keywords:  zzm321990 zzm321990 zzm321990 Rzzm321990 2zzm321990 *zzm321990 zzm321990 zzm321990 mapping; MRI; gradient recalled echo; self-supervised deep learning

Mesh:

Year:  2020        PMID: 32767489     DOI: 10.1002/mrm.28344

Source DB:  PubMed          Journal:  Magn Reson Med        ISSN: 0740-3194            Impact factor:   4.668


  4 in total

1.  Self-supervised IVIM DWI parameter estimation with a physics based forward model.

Authors:  Serge Didenko Vasylechko; Simon K Warfield; Onur Afacan; Sila Kurugol
Journal:  Magn Reson Med       Date:  2021-10-22       Impact factor: 4.668

2.  Deformation-Compensated Learning for Image Reconstruction Without Ground Truth.

Authors:  Weijie Gan; Yu Sun; Cihat Eldeniz; Jiaming Liu; Hongyu An; Ulugbek S Kamilov
Journal:  IEEE Trans Med Imaging       Date:  2022-08-31       Impact factor: 11.037

3.  Quantitative Gradient Echo MRI Identifies Dark Matter as a New Imaging Biomarker of Neurodegeneration that Precedes Tisssue Atrophy in Early Alzheimer's Disease.

Authors:  Satya V V N Kothapalli; Tammie L Benzinger; Andrew J Aschenbrenner; Richard J Perrin; Charles F Hildebolt; Manu S Goyal; Anne M Fagan; Marcus E Raichle; John C Morris; Dmitriy A Yablonskiy
Journal:  J Alzheimers Dis       Date:  2022       Impact factor: 4.472

4.  MRI-Based Classification of Neuropsychiatric Systemic Lupus Erythematosus Patients With Self-Supervised Contrastive Learning.

Authors:  Francesca Inglese; Minseon Kim; Gerda M Steup-Beekman; Tom W J Huizinga; Mark A van Buchem; Jeroen de Bresser; Dae-Shik Kim; Itamar Ronen
Journal:  Front Neurosci       Date:  2022-02-16       Impact factor: 4.677

  4 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.