Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches
Summary: Background: Early diagnosis of major depressive disorder (MDD) could enable timely interventions and effective management which subsequently improve clinical outcomes. However, quantitative and objective assessment tools for the suspected cases who present with depressive symptoms have not...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-05-01
|
| Series: | EBioMedicine |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352396422002110 |
| _version_ | 1811302821993119744 |
|---|---|
| author | Zhifei Li Roger S. McIntyre Syeda F. Husain Roger Ho Bach X. Tran Hien Thu Nguyen Shuenn-Chiang Soo Cyrus S. Ho Nanguang Chen |
| author_facet | Zhifei Li Roger S. McIntyre Syeda F. Husain Roger Ho Bach X. Tran Hien Thu Nguyen Shuenn-Chiang Soo Cyrus S. Ho Nanguang Chen |
| author_sort | Zhifei Li |
| collection | DOAJ |
| description | Summary: Background: Early diagnosis of major depressive disorder (MDD) could enable timely interventions and effective management which subsequently improve clinical outcomes. However, quantitative and objective assessment tools for the suspected cases who present with depressive symptoms have not been fully established. Methods: Based on a large-scale dataset (n = 363 subjects) collected with functional near-infrared spectroscopy (fNIRS) measurements during the verbal fluency task (VFT), this study proposed a data representation method for extracting spatiotemporal characteristics of NIRS signals, which emerged as candidate predictors in a two-phase machine learning framework to detect distinctive biomarkers for MDD. Supervised classifiers (e.g., support vector machine (SVM), k-nearest neighbors (KNN)) cooperated with cross-validation were implemented to evaluate the predictive capability of selected features in a training set. Another test set that was not involved in developing the algorithms enabled the independent assessment of the model's generalization. Findings: For the classification with the optimal fusion features, the SVM classifier achieved the highest accuracy of 75.6% ± 4.7% in the nested cross-validation, and the correct prediction rate of 78.0% with a sensitivity of 75.0% and a specificity of 81.4% in the test set. Moreover, the multiway ANOVA test on clinical and demographic factors confirmed that twenty out of 39 optimal features were significantly correlated with the MDD-distinctive consequence. Interpretation: The abnormal prefrontal activity of MDD may be quantified as diminished relative intensity and inappropriate activation timing of hemodynamic response, resulting in an objectively measurable biomarker for assessing cognitive deficits and screening MDD at the early stage. Funding: This study was funded by NUS iHeathtech Other Operating Expenses (R-722-000-004-731). |
| first_indexed | 2024-04-13T07:36:12Z |
| format | Article |
| id | doaj.art-c29d25fa84ba4f78ac9d949a9160cc32 |
| institution | Directory Open Access Journal |
| issn | 2352-3964 |
| language | English |
| last_indexed | 2024-04-13T07:36:12Z |
| publishDate | 2022-05-01 |
| publisher | Elsevier |
| record_format | Article |
| series | EBioMedicine |
| spelling | doaj.art-c29d25fa84ba4f78ac9d949a9160cc322022-12-22T02:56:07ZengElsevierEBioMedicine2352-39642022-05-0179104027Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approachesZhifei Li0Roger S. McIntyre1Syeda F. Husain2Roger Ho3Bach X. Tran4Hien Thu Nguyen5Shuenn-Chiang Soo6Cyrus S. Ho7Nanguang Chen8Institute of Health Innovation and Technology (iHealthtech), National University of Singapore, Singapore; Department of Biomedical Engineering, Faculty of Engineering, National University of Singapore, SingaporeMood Disorders Psychopharmacology Unit, University Health Network, University of Toronto, Toronto, ON, Canada; Canadian Rapid Treatment Center of Excellence, Mississauga, ON, CanadaDepartment of Paediatrics, Yong Loo Lin School of Medicine, NUS, SingaporeInstitute of Health Innovation and Technology (iHealthtech), National University of Singapore, Singapore; Department of Psychological Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore; Corresponding author at: Department of Psychological Medicine, Yong Loo Lin School of Medicine L9 NUHS Tower Block, 1E Kent Ridge Road, National University of Singapore.Institute for Preventive Medicine and Public Health, Hanoi Medical University, Viet Nam; Bloomberg School of Public Health, Johns Hopkins University, USAInstitute for Global Health Innovations, Duy Tan University, Viet Nam; Faculty of Medicine, Duy Tan University, Da Nang, Viet NamDepartment of Psychological Medicine, National University Hospital, SingaporeDepartment of Psychological Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore; Department of Psychological Medicine, National University Hospital, SingaporeInstitute of Health Innovation and Technology (iHealthtech), National University of Singapore, Singapore; Department of Biomedical Engineering, Faculty of Engineering, National University of Singapore, Singapore; NUS (Suzhou) Research Institute, Suzhou, ChinaSummary: Background: Early diagnosis of major depressive disorder (MDD) could enable timely interventions and effective management which subsequently improve clinical outcomes. However, quantitative and objective assessment tools for the suspected cases who present with depressive symptoms have not been fully established. Methods: Based on a large-scale dataset (n = 363 subjects) collected with functional near-infrared spectroscopy (fNIRS) measurements during the verbal fluency task (VFT), this study proposed a data representation method for extracting spatiotemporal characteristics of NIRS signals, which emerged as candidate predictors in a two-phase machine learning framework to detect distinctive biomarkers for MDD. Supervised classifiers (e.g., support vector machine (SVM), k-nearest neighbors (KNN)) cooperated with cross-validation were implemented to evaluate the predictive capability of selected features in a training set. Another test set that was not involved in developing the algorithms enabled the independent assessment of the model's generalization. Findings: For the classification with the optimal fusion features, the SVM classifier achieved the highest accuracy of 75.6% ± 4.7% in the nested cross-validation, and the correct prediction rate of 78.0% with a sensitivity of 75.0% and a specificity of 81.4% in the test set. Moreover, the multiway ANOVA test on clinical and demographic factors confirmed that twenty out of 39 optimal features were significantly correlated with the MDD-distinctive consequence. Interpretation: The abnormal prefrontal activity of MDD may be quantified as diminished relative intensity and inappropriate activation timing of hemodynamic response, resulting in an objectively measurable biomarker for assessing cognitive deficits and screening MDD at the early stage. Funding: This study was funded by NUS iHeathtech Other Operating Expenses (R-722-000-004-731).http://www.sciencedirect.com/science/article/pii/S2352396422002110Functional near-infrared spectroscopyDepressive disorderFeature selectionSupervised learningBiomarkers discoveryDepression |
| spellingShingle | Zhifei Li Roger S. McIntyre Syeda F. Husain Roger Ho Bach X. Tran Hien Thu Nguyen Shuenn-Chiang Soo Cyrus S. Ho Nanguang Chen Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches EBioMedicine Functional near-infrared spectroscopy Depressive disorder Feature selection Supervised learning Biomarkers discovery Depression |
| title | Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches |
| title_full | Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches |
| title_fullStr | Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches |
| title_full_unstemmed | Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches |
| title_short | Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches |
| title_sort | identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches |
| topic | Functional near-infrared spectroscopy Depressive disorder Feature selection Supervised learning Biomarkers discovery Depression |
| url | http://www.sciencedirect.com/science/article/pii/S2352396422002110 |
| work_keys_str_mv | AT zhifeili identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT rogersmcintyre identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT syedafhusain identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT rogerho identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT bachxtran identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT hienthunguyen identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT shuennchiangsoo identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT cyrussho identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches AT nanguangchen identifyingneuroimagingbiomarkersofmajordepressivedisorderfromcorticalhemodynamicresponsesusingmachinelearningapproaches |