Simultaneous Utilization of Mood Disorder Questionnaire and Bipolar Spectrum Diagnostic Scale for Machine Learning-Based Classification of Patients With Bipolar Disorders and Depressive Disorders

• DOI: 10.30773/pi.2023.0361
• Publication Date: 2024-08-01T05:55:33Z
• Authors: Kyungwon Kim; Hyun Ju Lim; Je-Min Park; Byung-Dae Lee; Young-Min Lee; Hwagyu Suh; Eunsoo Moon
• Keywords: 03 medical and health sciences; 0302 clinical medicine; Original Article

Objective Bipolar and depressive disorders are distinct with clearly different clinical courses, however, distinguishing between them often presents challenges. This study investigates the utility of self-report questionnaires, Mood Disorder Questionnaire (MDQ) Spectrum Diagnostic Scale (BSDS), machine learning-based multivariate analysis, to classify patients bipolar disorders.Methods A total 189 were included in study, all participants completed both MDQ BSDS questionnaires. Machine-learning classifiers, including support vector (SVM) linear discriminant analysis (LDA), exploited for analysis. Classification performance was assessed through cross-validation.Results Both demonstrated significant differences each item scores two groups. Machine SVM, achieved excellent discrimination levels area under ROC curve (AUC) values exceeding 0.8 questionnaire individually. In particular, combination further improved classification performance, yielding an AUC 0.8762.Conclusion suggests application learning can assist disorders. The potential combining high-dimensional psychiatric data as effective approach