msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping

• Published in: NeuroImage (Orlando, Fla.) Vol. 275; p. 120181
• Main Authors: He, Junjie, Peng, Yunsong, Fu, Bangkang, Zhu, Yuemin, Wang, Lihui, Wang, Rongpin
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 15.07.2023; Elsevier Limited; Elsevier
• Subjects: Algorithms; Alzheimer Disease - diagnostic imaging; Alzheimer's disease; Arbitrary resolution; Brain - diagnostic imaging; Brain Mapping - methods; Computer Science; Deep Learning; Differential diagnosis; Fourier transforms; Humans; Image Processing, Computer-Assisted - methods; Life Sciences; Machine Learning; Magnetic fields; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Magnetic susceptibility; Mapping; Medical Imaging; Methods; Morphology; Movement disorders; Neostriatum; Neurodegenerative Diseases; Neurons and Cognition; Parkinson Disease; Parkinson's disease; Self-supervised learning; Substantia nigra; Susceptibility; Susceptibility quantitative mapping; Arbitrary resolution; Susceptibility quantitative mapping; Alzheimer’s disease; Parkinson’s disease; Self-supervised learning; self-supervised learning; susceptibility quantitative mapping; morphological loss; arbitrary resolution
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2023.120181

•Self-supervised learning for arbitrary resolution quantitative susceptibility mapping (QSM), trained on one resolution data.•Morphology-based loss to reduce artifacts effectively and save training time efficiently.•Morphological QSM builder for decoupling dependence of the QSM on resolution with apriori information.•Significant changes in magnetic susceptibility in the brain regions of patients in the progression of AD or with PD. Quantitative susceptibility mapping (QSM) has been applied to the measurement of iron deposition and the auxiliary diagnosis of neurodegenerative disease. There still exists a dipole inversion problem in QSM reconstruction. Recently, deep learning approaches have been proposed to resolve this problem. However, most of these approaches are supervised methods that need pairs of the input phase and ground-truth. It remains a challenge to train a model for all resolutions without using the ground-truth and only using one resolution data. To address this, we proposed a self-supervised QSM deep learning method based on morphology. It consists of a morphological QSM builder to decouple the dependency of the QSM on acquisition resolution, and a morphological loss to reduce artifacts effectively and save training time efficiently. The proposed method can reconstruct arbitrary resolution QSM on both human data and animal data, regardless of whether the resolution is higher or lower than that of the training set. Our method outperforms the previous best unsupervised method with a 3.6% higher peak signal-to-noise ratio, 16.2% lower normalized root mean square error, and 22.1% lower high-frequency error norm. The morphological loss reduces training time by 22.1% with respect to the cycle gradient loss used in the previous unsupervised methods. Experimental results show that the proposed method accurately measures QSM with arbitrary resolutions, and achieves state-of-the-art results among unsupervised deep learning methods. Research on applications in neurodegenerative diseases found that our method is robust enough to measure significant increase in striatal magnetic susceptibility in patients during Alzheimer’s disease progression, as well as significant increase in substantia nigra susceptibility in Parkinson’s disease patients, and can be used as an auxiliary differential diagnosis tool for Alzheimer’s disease and Parkinson’s disease.