STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data

<b>Background</b>: Early diagnosis of depression is crucial for effective treatment and suicide prevention. Traditional methods rely on self-report questionnaires and clinical assessments, lacking objective biomarkers. Combining functional magnetic resonance imaging (fMRI) with artificia...

Full description

Saved in:
Bibliographic Details
Main Authors: Wei Zhang, Weiming Zeng, Hongyu Chen, Jie Liu, Hongjie Yan, Kaile Zhang, Ran Tao, Wai Ting Siok, Nizhuan Wang
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Tomography
Subjects:
Online Access:https://www.mdpi.com/2379-139X/10/12/138
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850085627851177984
author Wei Zhang
Weiming Zeng
Hongyu Chen
Jie Liu
Hongjie Yan
Kaile Zhang
Ran Tao
Wai Ting Siok
Nizhuan Wang
author_facet Wei Zhang
Weiming Zeng
Hongyu Chen
Jie Liu
Hongjie Yan
Kaile Zhang
Ran Tao
Wai Ting Siok
Nizhuan Wang
author_sort Wei Zhang
collection DOAJ
description <b>Background</b>: Early diagnosis of depression is crucial for effective treatment and suicide prevention. Traditional methods rely on self-report questionnaires and clinical assessments, lacking objective biomarkers. Combining functional magnetic resonance imaging (fMRI) with artificial intelligence can enhance depression diagnosis using neuroimaging indicators, but depression-specific fMRI datasets are often small and imbalanced, posing challenges for classification models. <b>New Method</b>: We propose the Spatio-Temporal Aggregation Network (STANet) for diagnosing depression by integrating convolutional neural networks (CNN) and recurrent neural networks (RNN) to capture both temporal and spatial features of brain activity. STANet comprises the following steps: (1) Aggregate spatio-temporal information via independent component analysis (ICA). (2) Utilize multi-scale deep convolution to capture detailed features. (3) Balance data using the synthetic minority over-sampling technique (SMOTE) to generate new samples for minority classes. (4) Employ the attention-Fourier gate recurrent unit (AFGRU) classifier to capture long-term dependencies, with an adaptive weight assignment mechanism to enhance model generalization. <b>Results</b>: STANet achieves superior depression diagnostic performance, with 82.38% accuracy and a 90.72% AUC. The Spatio-Temporal Feature Aggregation module enhances classification by capturing deeper features at multiple scales. The AFGRU classifier, with adaptive weights and a stacked Gated Recurrent Unit (GRU), attains higher accuracy and AUC. SMOTE outperforms other oversampling methods. Additionally, spatio-temporal aggregated features achieve better performance compared to using only temporal or spatial features. <b>Comparison with existing methods</b>: STANet significantly outperforms traditional classifiers, deep learning classifiers, and functional connectivity-based classifiers. <b>Conclusions</b>: The successful performance of STANet contributes to enhancing the diagnosis and treatment assessment of depression in clinical settings on imbalanced and small fMRI.
format Article
id doaj-art-e5083feb3b514c12b1756e6a4e578659
institution DOAJ
issn 2379-1381
2379-139X
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Tomography
spelling doaj-art-e5083feb3b514c12b1756e6a4e5786592025-08-20T02:43:41ZengMDPI AGTomography2379-13812379-139X2024-11-0110121895191410.3390/tomography10120138STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI DataWei Zhang0Weiming Zeng1Hongyu Chen2Jie Liu3Hongjie Yan4Kaile Zhang5Ran Tao6Wai Ting Siok7Nizhuan Wang8Lab of Digital Image and Intelligent Computation, College of Information Engineering, Shanghai Maritime University, Shanghai 201306, ChinaLab of Digital Image and Intelligent Computation, College of Information Engineering, Shanghai Maritime University, Shanghai 201306, ChinaLab of Digital Image and Intelligent Computation, College of Information Engineering, Shanghai Maritime University, Shanghai 201306, ChinaLab of Digital Image and Intelligent Computation, College of Information Engineering, Shanghai Maritime University, Shanghai 201306, ChinaDepartment of Neurology, Affiliated Lianyungang Hospital of Xuzhou Medical University, Lianyungang 222002, ChinaDepartment of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hong Kong, ChinaDepartment of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hong Kong, ChinaDepartment of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hong Kong, ChinaDepartment of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hong Kong, China<b>Background</b>: Early diagnosis of depression is crucial for effective treatment and suicide prevention. Traditional methods rely on self-report questionnaires and clinical assessments, lacking objective biomarkers. Combining functional magnetic resonance imaging (fMRI) with artificial intelligence can enhance depression diagnosis using neuroimaging indicators, but depression-specific fMRI datasets are often small and imbalanced, posing challenges for classification models. <b>New Method</b>: We propose the Spatio-Temporal Aggregation Network (STANet) for diagnosing depression by integrating convolutional neural networks (CNN) and recurrent neural networks (RNN) to capture both temporal and spatial features of brain activity. STANet comprises the following steps: (1) Aggregate spatio-temporal information via independent component analysis (ICA). (2) Utilize multi-scale deep convolution to capture detailed features. (3) Balance data using the synthetic minority over-sampling technique (SMOTE) to generate new samples for minority classes. (4) Employ the attention-Fourier gate recurrent unit (AFGRU) classifier to capture long-term dependencies, with an adaptive weight assignment mechanism to enhance model generalization. <b>Results</b>: STANet achieves superior depression diagnostic performance, with 82.38% accuracy and a 90.72% AUC. The Spatio-Temporal Feature Aggregation module enhances classification by capturing deeper features at multiple scales. The AFGRU classifier, with adaptive weights and a stacked Gated Recurrent Unit (GRU), attains higher accuracy and AUC. SMOTE outperforms other oversampling methods. Additionally, spatio-temporal aggregated features achieve better performance compared to using only temporal or spatial features. <b>Comparison with existing methods</b>: STANet significantly outperforms traditional classifiers, deep learning classifiers, and functional connectivity-based classifiers. <b>Conclusions</b>: The successful performance of STANet contributes to enhancing the diagnosis and treatment assessment of depression in clinical settings on imbalanced and small fMRI.https://www.mdpi.com/2379-139X/10/12/138depressionfMRIindependent component analysis (ICA)GRUsynthetic minority over-sampling technique (SMOTE)adaptive fusion weight
spellingShingle Wei Zhang
Weiming Zeng
Hongyu Chen
Jie Liu
Hongjie Yan
Kaile Zhang
Ran Tao
Wai Ting Siok
Nizhuan Wang
STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data
Tomography
depression
fMRI
independent component analysis (ICA)
GRU
synthetic minority over-sampling technique (SMOTE)
adaptive fusion weight
title STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data
title_full STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data
title_fullStr STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data
title_full_unstemmed STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data
title_short STANet: A Novel Spatio-Temporal Aggregation Network for Depression Classification with Small and Unbalanced FMRI Data
title_sort stanet a novel spatio temporal aggregation network for depression classification with small and unbalanced fmri data
topic depression
fMRI
independent component analysis (ICA)
GRU
synthetic minority over-sampling technique (SMOTE)
adaptive fusion weight
url https://www.mdpi.com/2379-139X/10/12/138
work_keys_str_mv AT weizhang stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT weimingzeng stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT hongyuchen stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT jieliu stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT hongjieyan stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT kailezhang stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT rantao stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT waitingsiok stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata
AT nizhuanwang stanetanovelspatiotemporalaggregationnetworkfordepressionclassificationwithsmallandunbalancedfmridata