Elif Sarisik, David Popovic, Daniel Keeser, Adyasha Khuntia, Kolja Schiltz, Peter Falkai, Oliver Pogarell, Nikolaos Koutsouleris
{"title":"EEG-based Signatures of Schizophrenia, Depression, and Aberrant Aging: A Supervised Machine Learning Investigation.","authors":"Elif Sarisik, David Popovic, Daniel Keeser, Adyasha Khuntia, Kolja Schiltz, Peter Falkai, Oliver Pogarell, Nikolaos Koutsouleris","doi":"10.1093/schbul/sbae150","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Electroencephalography (EEG) is a noninvasive, cost-effective, and robust tool, which directly measures in vivo neuronal mass activity with high temporal resolution. Combined with state-of-the-art machine learning (ML) techniques, EEG recordings could potentially yield in silico biomarkers of severe mental disorders.</p><p><strong>Hypothesis: </strong>Pathological and physiological aging processes influence the electrophysiological signatures of schizophrenia (SCZ) and major depressive disorder (MDD).</p><p><strong>Study design: </strong>From a single-center cohort (N = 735, 51.6% male) comprising healthy control individuals (HC, N = 245) and inpatients suffering from SCZ (N = 250) or MDD (N = 240), we acquired resting-state 19 channel-EEG recordings. Using repeated nested cross-validation, support vector machine models were trained to (1) classify patients with SCZ or MDD and HC individuals and (2) predict age in HC individuals. The age model was applied to patient groups to calculate Electrophysiological Age Gap Estimation (EphysAGE) as the difference between predicted and chronological age. The links between EphysAGE, diagnosis, and medication were then further explored.</p><p><strong>Study results: </strong>The classification models robustly discriminated SCZ from HC (balanced accuracy, BAC = 72.7%, P < .001), MDD from HC (BAC = 67.0%, P < .001), and SCZ from MDD individuals (BAC = 63.2%, P < .001). Notably, central alpha (8-11 Hz) power decrease was the most consistently predictive feature for SCZ and MDD. Higher EphysAGE was associated with an increased likelihood of being misclassified as SCZ in HC and MDD (ρHC = 0.23, P < .001; ρMDD = 0.17, P = .01).</p><p><strong>Conclusions: </strong>ML models can extract electrophysiological signatures of MDD and SCZ for potential clinical use. However, the impact of aging processes on diagnostic separability calls for timely application of such models, possibly in early recognition settings.</p>","PeriodicalId":21530,"journal":{"name":"Schizophrenia Bulletin","volume":" ","pages":""},"PeriodicalIF":5.3000,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Schizophrenia Bulletin","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1093/schbul/sbae150","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PSYCHIATRY","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Electroencephalography (EEG) is a noninvasive, cost-effective, and robust tool, which directly measures in vivo neuronal mass activity with high temporal resolution. Combined with state-of-the-art machine learning (ML) techniques, EEG recordings could potentially yield in silico biomarkers of severe mental disorders.
Hypothesis: Pathological and physiological aging processes influence the electrophysiological signatures of schizophrenia (SCZ) and major depressive disorder (MDD).
Study design: From a single-center cohort (N = 735, 51.6% male) comprising healthy control individuals (HC, N = 245) and inpatients suffering from SCZ (N = 250) or MDD (N = 240), we acquired resting-state 19 channel-EEG recordings. Using repeated nested cross-validation, support vector machine models were trained to (1) classify patients with SCZ or MDD and HC individuals and (2) predict age in HC individuals. The age model was applied to patient groups to calculate Electrophysiological Age Gap Estimation (EphysAGE) as the difference between predicted and chronological age. The links between EphysAGE, diagnosis, and medication were then further explored.
Study results: The classification models robustly discriminated SCZ from HC (balanced accuracy, BAC = 72.7%, P < .001), MDD from HC (BAC = 67.0%, P < .001), and SCZ from MDD individuals (BAC = 63.2%, P < .001). Notably, central alpha (8-11 Hz) power decrease was the most consistently predictive feature for SCZ and MDD. Higher EphysAGE was associated with an increased likelihood of being misclassified as SCZ in HC and MDD (ρHC = 0.23, P < .001; ρMDD = 0.17, P = .01).
Conclusions: ML models can extract electrophysiological signatures of MDD and SCZ for potential clinical use. However, the impact of aging processes on diagnostic separability calls for timely application of such models, possibly in early recognition settings.
期刊介绍:
Schizophrenia Bulletin seeks to review recent developments and empirically based hypotheses regarding the etiology and treatment of schizophrenia. We view the field as broad and deep, and will publish new knowledge ranging from the molecular basis to social and cultural factors. We will give new emphasis to translational reports which simultaneously highlight basic neurobiological mechanisms and clinical manifestations. Some of the Bulletin content is invited as special features or manuscripts organized as a theme by special guest editors. Most pages of the Bulletin are devoted to unsolicited manuscripts of high quality that report original data or where we can provide a special venue for a major study or workshop report. Supplement issues are sometimes provided for manuscripts reporting from a recent conference.