Fred Hutch AI Club - July 2026 Talk
Dr. Nida Itrat Abbasi will discuss the development of responsible AI frameworks for mental wellbeing assessment and offer broader insights into the evaluation of AI systems deployed in sensitive human-centred domains.
Abstract:
The increasing need for accessible mental wellbeing support has driven the development of AI-powered approaches that can complement traditional psychological assessment methods. While self-report questionnaires remain the clinical standard, they often overlook the rich behavioural cues that accompany emotional wellbeing, particularly in children. Recent advances in multimodal artificial intelligence present an opportunity to incorporate verbal and non-verbal behavioural information into more comprehensive assessment frameworks.
This talk presents research from Dr. Nida Itrat Abbasi’s doctoral work on AI-powered mental wellbeing assessment within robot-driven interactions with children. Using multimodal behavioural data like speech, facial expressions and language collected during child-robot interactions, she examined how AI models can be leveraged to infer mental wellbeing. Beyond evaluating predictive performance, the work investigates how do AI models make their decisions. The findings contribute to the development of responsible AI frameworks for mental wellbeing assessment and offer broader insights into the evaluation of AI systems deployed in sensitive human-centred domains.
Free