USC Researchers Combine Multimodal Signals with AI to Detect Depression and Suicide Risk
Researchers at the University of Southern California's Viterbi School of Engineering have published findings on a novel approach to detecting depression and suicide risk using a combination of physiological measurements and artificial intelligence.
The study integrates three distinct types of biosignals: electrodermal activity (sweat), brain signals (likely via EEG), and eye movement data. Each of these biomarkers provides different insights into the body's stress response and neurological state, which are known to correlate with mental health conditions.
By combining these multiple signal types with machine learning algorithms, the research team aimed to create a more robust detection system than could be achieved with a single measurement modality. Multimodal approaches in AI research have shown promise in various medical diagnostic applications because they can capture complementary aspects of a condition that a single data source might miss.
The study contributes to an expanding area of research focused on objective, technology-assisted mental health assessment. Traditional diagnosis of depression and evaluation of suicide risk rely heavily on clinical interviews and standardized questionnaires, which can be subject to variability. AI-based approaches using physiological markers offer the potential for more consistent and continuous monitoring.
The findings were published by USC Viterbi, whose researchers have been working at the intersection of engineering and healthcare to develop technologies for early detection of mental health conditions.