msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping
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 proble...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-07-01
|
| Series: | NeuroImage |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1053811923003324 |
| _version_ | 1827934007638097920 |
|---|---|
| author | Junjie He Yunsong Peng Bangkang Fu Yuemin Zhu Lihui Wang Rongpin Wang |
| author_facet | Junjie He Yunsong Peng Bangkang Fu Yuemin Zhu Lihui Wang Rongpin Wang |
| author_sort | Junjie He |
| collection | DOAJ |
| description | 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. |
| first_indexed | 2024-03-13T07:33:23Z |
| format | Article |
| id | doaj.art-6a2e2c244b524b7f9d10cc807ed69efe |
| institution | Directory Open Access Journal |
| issn | 1095-9572 |
| language | English |
| last_indexed | 2024-03-13T07:33:23Z |
| publishDate | 2023-07-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage |
| spelling | doaj.art-6a2e2c244b524b7f9d10cc807ed69efe2023-06-04T04:23:15ZengElsevierNeuroImage1095-95722023-07-01275120181msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mappingJunjie He0Yunsong Peng1Bangkang Fu2Yuemin Zhu3Lihui Wang4Rongpin Wang5Engineering Research Center of Text Computing & Cognitive Intelligence, Ministry of Education, Key Laboratory of Intelligent Medical Image Analysis and Precise Diagnosis of Guizhou Province, State Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University, No. 2870, Huaxi Avenue South, Guiyang 550025, Guizhou, China; Department of Radiology, International Exemplary Cooperation Base of Precision Imaging for Diagnosis and Treatment, Guizhou Provincial People’s Hospital, No. 83, Zhongshan Dong Road, Guiyang, 550002, Guizhou, ChinaDepartment of Radiology, International Exemplary Cooperation Base of Precision Imaging for Diagnosis and Treatment, Guizhou Provincial People’s Hospital, No. 83, Zhongshan Dong Road, Guiyang, 550002, Guizhou, ChinaDepartment of Radiology, International Exemplary Cooperation Base of Precision Imaging for Diagnosis and Treatment, Guizhou Provincial People’s Hospital, No. 83, Zhongshan Dong Road, Guiyang, 550002, Guizhou, ChinaCREATIS, IRP Metislab, University of Lyon, INSA Lyon, CNRS UMR 5220, Inserm U1294, Lyon, FranceEngineering Research Center of Text Computing & Cognitive Intelligence, Ministry of Education, Key Laboratory of Intelligent Medical Image Analysis and Precise Diagnosis of Guizhou Province, State Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University, No. 2870, Huaxi Avenue South, Guiyang 550025, Guizhou, ChinaCorresponding author.; Department of Radiology, International Exemplary Cooperation Base of Precision Imaging for Diagnosis and Treatment, Guizhou Provincial People’s Hospital, No. 83, Zhongshan Dong Road, Guiyang, 550002, Guizhou, ChinaQuantitative 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.http://www.sciencedirect.com/science/article/pii/S1053811923003324Susceptibility quantitative mappingSelf-supervised learningArbitrary resolutionAlzheimer’s diseaseParkinson’s disease |
| spellingShingle | Junjie He Yunsong Peng Bangkang Fu Yuemin Zhu Lihui Wang Rongpin Wang msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping NeuroImage Susceptibility quantitative mapping Self-supervised learning Arbitrary resolution Alzheimer’s disease Parkinson’s disease |
| title | msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping |
| title_full | msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping |
| title_fullStr | msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping |
| title_full_unstemmed | msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping |
| title_short | msQSM: Morphology-based self-supervised deep learning for quantitative susceptibility mapping |
| title_sort | msqsm morphology based self supervised deep learning for quantitative susceptibility mapping |
| topic | Susceptibility quantitative mapping Self-supervised learning Arbitrary resolution Alzheimer’s disease Parkinson’s disease |
| url | http://www.sciencedirect.com/science/article/pii/S1053811923003324 |
| work_keys_str_mv | AT junjiehe msqsmmorphologybasedselfsuperviseddeeplearningforquantitativesusceptibilitymapping AT yunsongpeng msqsmmorphologybasedselfsuperviseddeeplearningforquantitativesusceptibilitymapping AT bangkangfu msqsmmorphologybasedselfsuperviseddeeplearningforquantitativesusceptibilitymapping AT yueminzhu msqsmmorphologybasedselfsuperviseddeeplearningforquantitativesusceptibilitymapping AT lihuiwang msqsmmorphologybasedselfsuperviseddeeplearningforquantitativesusceptibilitymapping AT rongpinwang msqsmmorphologybasedselfsuperviseddeeplearningforquantitativesusceptibilitymapping |