Voices for Mental Health: New funding for AI-powered depression detection

A2I2 researchers have received new national funding for Voices for Mental Health: AI‑Driven Diagnosis through Speech, Language, and other Data, a collaborative project led by psychiatrist Rudolf Uher with Frank Rudzicz and Sageev Oore leading the AI components.

The team will develop machine learning models that listen not just to what people say, but how they say it—capturing subtle patterns in tone, rhythm, word choice, and dialogue that are difficult for clinicians to track in real time. By combining these signals with clinical data, the project aims to support earlier and more accurate detection of conditions such as depression, bipolar disorder, and anxiety.

Crucially, the work is grounded in strong ethics and clinical partnerships. Youth and families will be involved throughout, helping to design tools that are transparent, fair, and actually useful in real-world care. The long‑term vision is a set of AI supports that can help clinicians monitor mental health over time, spot concerning changes sooner, and tailor treatment more precisely.

For A2I2 students, the project opens up rich training opportunities at the intersection of AI, psychiatry, and digital health—exactly the kind of interdisciplinary work our institute exists to enable.