Uncertainty-Aware and Lesion-Specific Image Synthesis in Multiple Sclerosis Magnetic Resonance Imaging: A Multicentric Validation Study

Tom Finck; Hongwei Li; Sarah Schlaeger; Lioba Grundl; Nico Sollmann; Benjamin Bender; Eva Bürkle; Claus Zimmer; Jan Kirschke; Björn Menze; Mark Mühlau; Benedikt Wiestler

doi:10.3389/fnins.2022.889808

Uncertainty-Aware and Lesion-Specific Image Synthesis in Multiple Sclerosis Magnetic Resonance Imaging: A Multicentric Validation Study

Front Neurosci. 2022 Apr 26:16:889808. doi: 10.3389/fnins.2022.889808. eCollection 2022.

Authors

Tom Finck¹, Hongwei Li², Sarah Schlaeger¹, Lioba Grundl¹, Nico Sollmann^{1

3}, Benjamin Bender⁴, Eva Bürkle⁴, Claus Zimmer¹, Jan Kirschke¹, Björn Menze², Mark Mühlau^{5

6}, Benedikt Wiestler^{1

2}

Affiliations

¹ Department of Diagnostic and Interventional Neuroradiology, School of Medicine, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany.
² Image-Based Biomedical Modeling, Technical University of Munich, Munich, Germany.
³ Department of Diagnostic and Interventional Radiology, University Hospital Ulm, Ulm, Germany.
⁴ Department of Diagnostic and Interventional Neuroradiology, Universitätsklinikum Tübingen, Tübingen, Germany.
⁵ TUM-Neuroimaging Center, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany.
⁶ Department of Neurology, School of Medicine, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany.

Abstract

Generative adversarial networks (GANs) can synthesize high-contrast MRI from lower-contrast input. Targeted translation of parenchymal lesions in multiple sclerosis (MS), as well as visualization of model confidence further augment their utility, provided that the GAN generalizes reliably across different scanners. We here investigate the generalizability of a refined GAN for synthesizing high-contrast double inversion recovery (DIR) images and propose the use of uncertainty maps to further enhance its clinical utility and trustworthiness. A GAN was trained to synthesize DIR from input fluid-attenuated inversion recovery (FLAIR) and T1w of 50 MS patients (training data). In another 50 patients (test data), two blinded readers (R1 and R2) independently quantified lesions in synthetic DIR (synthDIR), acquired DIR (trueDIR) and FLAIR. Of the 50 test patients, 20 were acquired on the same scanner as training data (internal data), while 30 were scanned at different scanners with heterogeneous field strengths and protocols (external data). Lesion-to-Background ratios (LBR) for MS-lesions vs. normal appearing white matter, as well as image quality parameters were calculated. Uncertainty maps were generated to visualize model confidence. Significantly more MS-specific lesions were found in synthDIR compared to FLAIR (R1: 26.7 ± 2.6 vs. 22.5 ± 2.2 p < 0.0001; R2: 22.8 ± 2.2 vs. 19.9 ± 2.0, p = 0.0005). While trueDIR remained superior to synthDIR in R1 [28.6 ± 2.9 vs. 26.7 ± 2.6 (p = 0.0021)], both sequences showed comparable lesion conspicuity in R2 [23.3 ± 2.4 vs. 22.8 ± 2.2 (p = 0.98)]. Importantly, improvements in lesion counts were similar in internal and external data. Measurements of LBR confirmed that lesion-focused GAN training significantly improved lesion conspicuity. The use of uncertainty maps furthermore helped discriminate between MS lesions and artifacts. In conclusion, this multicentric study confirms the external validity of a lesion-focused Deep-Learning tool aimed at MS imaging. When implemented, uncertainty maps are promising to increase the trustworthiness of synthetic MRI.

Keywords: artificial intelligence (AI); deep learning – artificial neural network (DL-ANN); double inversion recovery (DIR); magnetic resonance imaging; multiple sclerosis; neuroradiology; synthetic MRI.