An artificial intelligence-assisted screening framework called the Major Depressive Disorder Detection System, or MD3S, has been developed to identify individuals with documented major depressive disorder using routinely collected laboratory and clinical information. Reporting in Translational Psychiatry, the research team designed the computational framework to distinguish individuals diagnosed with major depressive disorder from non-depressed controls across multicenter cohorts.
Major depressive disorder represents a prevalent psychiatric condition that exerts substantial personal and societal impacts worldwide. Despite its significant health burden, the identification of affected individuals across large-scale populations continues to rely primarily on traditional symptom-based clinical assessment methods.
To construct and evaluate the screening framework in a retrospective multicenter setting, the investigators analyzed more than 180,000 real-world health assessment records drawn from the UK Biobank.
The MD3S framework operates by integrating routine hematological, biochemical, and physiological indicators captured during standard health evaluations. By analyzing these standard laboratory measurements, the model distinguishes individuals diagnosed with major depressive disorder from non-MDD control subjects.
During internal testing on the primary evaluation dataset, MD3S demonstrated strong screening capability, achieving an accuracy of 86.57 percent in classifying individuals with documented major depressive disorder and controls.
To assess whether the framework maintains cross-population generalizability beyond the initial development population, the researchers validated MD3S across two independent external hospital-based cohorts. The model maintained robust screening performance across both external clinical cohorts.
When evaluated against conventional machine learning algorithms and deep learning approaches, MD3S demonstrated improved predictive performance. The framework also exhibited stable transferability across different patient cohorts alongside high computational efficiency.
The study authors concluded that these findings highlight the potential for routine laboratory data to support scalable, artificial intelligence-assisted screening strategies. Such computational approaches could help identify individuals with major depressive disorder across diverse clinical environments and broader population settings.
Subject of Research: Artificial intelligence screening for major depressive disorder using routine clinical laboratory tests
Article Title: Artificial intelligence-assisted screening of major depressive disorder using routine laboratory tests across multicenter cohorts
Article References: Sha, Y., Pan, H., Luo, G., Meng, W., Xu, W., Feng, Z., Xing, A., Zheng, W., Zhai, X., Tong, H. H. Y., Lv, Y., Chen, X., Huang, M., & Li, K. (2026). Artificial intelligence-assisted screening of major depressive disorder using routine laboratory tests across multicenter cohorts. Translational Psychiatry. https://doi.org/10.1038/s41398-026-04413-8
Image Credits: AI Generated
DOI: 10.1038/s41398-026-04413-8
Keywords: Major Depressive Disorder, MD3S, Artificial Intelligence Screening, Routine Laboratory Tests, UK Biobank, Symptom-Based Clinical Assessment, Hematological Indicators, Biochemical Indicators, Physiological Indicators, Multicenter Cohorts, Internal Validation, External Hospital-Based Cohorts, Machine Learning, Deep Learning, Cross-Cohort Transferability, Translational Psychiatry
News Source: Blake Davidson. (November 16, 2026). AI System Screens for Major Depressive Disorder Using Routine Laboratory Tests. Scienmag.



