Subthreshold depression (StD) is a mild form of depression, not strong enough for a clinical diagnosis, but still a warning sign for future mental health issues. Think of it as being on the edge of full-blown depression.
People with depression often show fewer facial expressions, and past research has used facial cues to detect conditions like anxiety and stress. But here’s the catch: we still don’t know if StD affects how people express or recognize emotions on faces.
Scientists are now exploring whether subtle changes in facial behavior could help spot StD early, before it deepens into something more serious.
To explore how subtle facial cues might signal early signs of depression, Associate Professor Eriko Sugimori and doctoral student Mayu Yamaguchi at Waseda University studied Japanese undergraduates using facial data and artificial intelligence.
AI-powered phone app detects depression from facial expression
Their research uncovered distinct patterns of muscle movement linked to depressive symptoms, even in individuals who hadn’t been clinically diagnosed. These findings suggest that AI could help detect early emotional distress, offering a powerful tool for preventative mental health care.
Sugimori said, “As concerns around mental well-being have been rising, I wanted to explore how subtle non-verbal cues, such as facial expressions, shape social impressions and reflect mental health using artificial intelligence-based facial analysis.”
Researchers asked 64 Japanese students to film short self-intro videos. Another group of students watched these and rated how friendly, expressive, and natural the speakers seemed. Meanwhile, an AI tool called OpenFace 2.0 analyzed tiny facial muscle movements in the same videos.
The results showed a clear trend: students with mild depressive symptoms (StD) were seen as less friendly and expressive, but not stiff or fake. This means StD doesn’t make people look negative; it just softens their positive expressions.
Using AI to analyze student videos, researchers spotted subtle facial movements, as slight brow lifts, eye widening, and mouth stretches, that were more common in students with mild depressive symptoms (StD). These micro-expressions were strongly tied to depression scores, even though they were too subtle for most people to notice.
Notably, the study focused on Japanese students, where cultural norms shape how emotions are shown. So while the findings are powerful, they also highlight the need to consider cultural context when reading emotional cues.
Sugimori said, “Our novel approach of short self-introduction videos and automated facial expression analysis can be applied to screen and detect mental health in schools, universities, and workplaces.”
The proposed approach could be used in mental health technology, digital health platforms, or employee wellness programs to monitor psychological well-being efficiently.
“Overall, our study provides a novel, accessible, and non-invasive artificial intelligence-based facial analysis tool for early detection of depression (before the appearance of clinical symptoms), enabling early interventions and timely care of mental health,” concludes Sugimori.
Journal Reference:
- Sugimori, E., Yamaguchi, M. Subthreshold depression is associated with altered facial expression and impression formation via subjective ratings and action unit analysis. Sci Rep 15, 30761 (2025). DOI: 10.1038/s41598-025-15874-0