Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis
In this study, we propose a deep-learning technique for functional MRI analysis. We introduced a novel self-supervised learning scheme that is particularly useful for functional MRI wherein the subject identity is used as the teacher signal of a neural network. The neural network is trained solely b...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2021-08-01
|
Series: | Frontiers in Neuroscience |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2021.696853/full |
id |
doaj-9cb0870f5d114af8b728b0ac0857ba3d |
---|---|
record_format |
Article |
spelling |
doaj-9cb0870f5d114af8b728b0ac0857ba3d2021-09-03T10:55:48ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2021-08-011510.3389/fnins.2021.696853696853Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia DiagnosisYuki Hashimoto0Yousuke Ogata1Manabu Honda2Yuichi Yamashita3Department of Information Medicine, National Center of Neurology and Psychiatry, National Institute of Neuroscience, Kodaira, JapanInstitute of Innovative Research, Tokyo Institute of Technology, Yokohama, JapanDepartment of Information Medicine, National Center of Neurology and Psychiatry, National Institute of Neuroscience, Kodaira, JapanDepartment of Information Medicine, National Center of Neurology and Psychiatry, National Institute of Neuroscience, Kodaira, JapanIn this study, we propose a deep-learning technique for functional MRI analysis. We introduced a novel self-supervised learning scheme that is particularly useful for functional MRI wherein the subject identity is used as the teacher signal of a neural network. The neural network is trained solely based on functional MRI-scans, and the training does not require any explicit labels. The proposed method demonstrated that each temporal volume of resting state functional MRI contains enough information to identify the subject. The network learned a feature space in which the features were clustered per subject for the test data as well as for the training data; this is unlike the features extracted by conventional methods including region of interests (ROIs) pooling signals and principal component analysis. In addition, applying a simple linear classifier to the per-subject mean of the features (namely “identity feature”), we demonstrated that the extracted features could contribute to schizophrenia diagnosis. The classification accuracy of our identity features was comparable to that of the conventional functional connectivity. Our results suggested that our proposed training scheme of the neural network captured brain functioning related to the diagnosis of psychiatric disorders as well as the identity of the subject. Our results together highlight the validity of our proposed technique as a design for self-supervised learning.https://www.frontiersin.org/articles/10.3389/fnins.2021.696853/fulldeep-learningfunctional MRIneural networkfeature extractionpsychiatric diagnosisself-supervised learning |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Yuki Hashimoto Yousuke Ogata Manabu Honda Yuichi Yamashita |
spellingShingle |
Yuki Hashimoto Yousuke Ogata Manabu Honda Yuichi Yamashita Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis Frontiers in Neuroscience deep-learning functional MRI neural network feature extraction psychiatric diagnosis self-supervised learning |
author_facet |
Yuki Hashimoto Yousuke Ogata Manabu Honda Yuichi Yamashita |
author_sort |
Yuki Hashimoto |
title |
Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis |
title_short |
Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis |
title_full |
Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis |
title_fullStr |
Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis |
title_full_unstemmed |
Deep Feature Extraction for Resting-State Functional MRI by Self-Supervised Learning and Application to Schizophrenia Diagnosis |
title_sort |
deep feature extraction for resting-state functional mri by self-supervised learning and application to schizophrenia diagnosis |
publisher |
Frontiers Media S.A. |
series |
Frontiers in Neuroscience |
issn |
1662-453X |
publishDate |
2021-08-01 |
description |
In this study, we propose a deep-learning technique for functional MRI analysis. We introduced a novel self-supervised learning scheme that is particularly useful for functional MRI wherein the subject identity is used as the teacher signal of a neural network. The neural network is trained solely based on functional MRI-scans, and the training does not require any explicit labels. The proposed method demonstrated that each temporal volume of resting state functional MRI contains enough information to identify the subject. The network learned a feature space in which the features were clustered per subject for the test data as well as for the training data; this is unlike the features extracted by conventional methods including region of interests (ROIs) pooling signals and principal component analysis. In addition, applying a simple linear classifier to the per-subject mean of the features (namely “identity feature”), we demonstrated that the extracted features could contribute to schizophrenia diagnosis. The classification accuracy of our identity features was comparable to that of the conventional functional connectivity. Our results suggested that our proposed training scheme of the neural network captured brain functioning related to the diagnosis of psychiatric disorders as well as the identity of the subject. Our results together highlight the validity of our proposed technique as a design for self-supervised learning. |
topic |
deep-learning functional MRI neural network feature extraction psychiatric diagnosis self-supervised learning |
url |
https://www.frontiersin.org/articles/10.3389/fnins.2021.696853/full |
work_keys_str_mv |
AT yukihashimoto deepfeatureextractionforrestingstatefunctionalmribyselfsupervisedlearningandapplicationtoschizophreniadiagnosis AT yousukeogata deepfeatureextractionforrestingstatefunctionalmribyselfsupervisedlearningandapplicationtoschizophreniadiagnosis AT manabuhonda deepfeatureextractionforrestingstatefunctionalmribyselfsupervisedlearningandapplicationtoschizophreniadiagnosis AT yuichiyamashita deepfeatureextractionforrestingstatefunctionalmribyselfsupervisedlearningandapplicationtoschizophreniadiagnosis |
_version_ |
1717817459295125504 |