MHDeep
Research on detecting mental health disorders from smartwatch and smartphone sensor data using compact on-device neural networks.
Research
- Indication
- Mental health disorder monitoring and detection
- Data source
- Smartwatch and smartphone sensor data
- Collaborators
- Princeton University (research lineage)
- Partnership availability
- Seeking research and clinical partners
The problem
Mental illness affects hundreds of millions of people worldwide and is a leading cause of disability, yet stigma and limited access to services mean many conditions go undetected until they escalate.
Our approach
MHDeep explores whether mental health disorders can be detected from everyday sensor data. In a 2021 proof of concept, models built from eight categories of smartwatch and smartphone sensor data detected three disorders — schizoaffective disorder, major depressive disorder and bipolar disorder — with reported accuracies between 82.4% and 90.4% in the research setting.
Current status
Research. MHDeep is an early-stage research program; results to date are proof-of-concept findings, not clinical validation. This status does not imply regulatory clearance, approval or diagnostic status in any jurisdiction.
Collaboration need
We are seeking academic and clinical research partners to expand data collection and move toward prospective evaluation. Partner with us.
Evidence
- MHDeep — mental health disorder detection based on wearable sensors and neural networks (proof of concept, 2021) — Reported 82.4-90.4% accuracy across three disorders in the published research setting.
Program status reflects the stage stated above; it does not imply regulatory clearance, diagnostic availability, or commercial deployment unless explicitly stated.