Scottish Enterprise

SMART Grant 

Advancing Precision Medicine in Mental Health Screening: Employing Deep Learning AI Models to Detect Behavioural Markers of Psychiatric Symptoms

The current project seeks to further precision digital mental healthcare through the discovery of relevant markers in vocal and visual expression data to predict current state of mood and stress with the aim of measuring burnout risk. Vocal and visual expression data collection is much less onerous than eliciting self-reported symptom ratings, and so they are good candidates for ongoing mental health assessment that is not burdensome on individuals. Improved momentary symptom assessment will enable the next generation of targeted mental health care. In addition, there are many clinical settings in which accurate momentary assessment is critical to preventing harm.