Talking to machines: Al-Powered Speech Analysis for Mental Health; a MOHT x SYNAPXE Seminar and Networking Series

This talk given by our Chief Data Scientist, Dr Wilson Goh, will explore innovative Al techniques for mental health assessment through speech and emotion analysis. We present our research on using speech as a predictive biomarker for early detection of mental health disorders. Our study involves a various Al models trained on engineered sentiment and linguistic features from a high-risk cohort, highlighting the importance of dynamic emotional patterns. We will discuss preliminary findings on how large language models can be taught to differentiate risk using emotion information. Finally, we emphasise the need for interdisciplinary collaboration to develop effective Al-driven tools for mental health care in primary and community settings