Developing Machine Learning Multi-Model Screening Tools for Mental Health Using Voice, Video, and Eye-Tracking Data.

This research aims to clinically validate a novel artificial intelligence based tool designed to screen for mental wellness indicators such as stress, anxiety, depression, burnout, and resilience. The tool leverages non-invasive multimodal data including voice patterns, facial microexpressions, and eye movements to assess psychological well-being. The primary objective is to develop an explainable, culturally relevant, and clinically reliable digital screening model specifically tailored to the Indian population. This AI tool is envisioned to facilitate early identification and timely intervention in mental health care by offering an accessible, scalable, and user-friendly technological solution. The study will adopt a cross-sectional observational design to collect and analyze multimodal biometric data from consenting participants aged between 14 to 65 years. Data collection will involve non-invasive audio and video recordings during guided psychological tasks, alongside psychological assessments using gold-standard questionnaires and clinical observations. It is important to note that the model is intended as a screening and well-being support tool and is not a substitute for clinical diagnosis.