Generalizable synthetic MRI with physics-informed convolutional networks

Abstract

In this study, we develop a physics-informed deep learning-based method to synthesize multiple brain magnetic resonance imaging (MRI) contrasts from a single five-minute acquisition and investigate its ability to generalize to arbitrary contrasts to accelerate neuroimaging protocols. A dataset of fifty-five subjects acquired with a standard MRI protocol and a five-minute transient-state sequence was used to develop a physics-informed deep learning-based method. The model, based on a generative adversarial network, maps data acquired from the five-minute scan to "effective" quantitative parameter maps, here named q*-maps, by using its generated PD, T1, and T2 values in a signal model to synthesize four standard contrasts (proton density-weighted, T1-weighted, T2-weighted, and T2-weighted fluid-attenuated inversion recovery), from which losses are computed. The q*-maps are compared to literature values and the synthetic contrasts are compared to an end-to-end deep learning-based method proposed by literature. The generalizability of the proposed method is investigated for five volunteers by synthesizing three non-standard contrasts unseen during training and comparing these to respective ground truth acquisitions via contrast-to-noise ratio and quantitative assessment. The physics-informed method was able to match the high-quality synthMRI of the end-to-end method for the four standard contrasts, with mean \pm standard deviation structural similarity metrics above 0.75 \pm 0.08 and peak signal-to-noise ratios above 22.4 \pm 1.9 and 22.6 \pm 2.1. Additionally, the physics-informed method provided retrospective contrast adjustment, with visually similar signal contrast and comparable contrast-to-noise ratios to the ground truth acquisitions for three sequences unused for model training, demonstrating its generalizability and potential application to accelerate neuroimaging protocols.

arXiv comment23 pages, 7 figures, 1 table. Presented at ISMRM 2022. Will be submitted to NMR in biomedicine

References in corpus (10)

• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Road Extraction by Deep Residual U-Net
• An overview of deep learning in medical imaging focusing on MRI
• Generative Adversarial Network in Medical Imaging: A Review
• Deep learning-based synthetic-CT generation in radiotherapy and PET: a review
• Fast quantitative MRI as a nonlinear tomography problem
• High resolution in-vivo MR-STAT using a matrix-free and parallelized reconstruction algorithm
• A three-dimensional MR-STAT protocol for high-resolution multi-parametric quantitative MRI
• High-fidelity Direct Contrast Synthesis from Magnetic Resonance Fingerprinting

Cited by in corpus (1)

• Quantitative mapping from conventional MRI using self-supervised physics-guided deep learning: applications to a large-scale, clinically heterogeneous dataset