Sensing the Mind: Human-Centric AI for Smart Mental Healthcare

Join us in exploring how AI-driven technologies can transform the detection, monitoring, and support of mental health in an increasingly complex world

Mental health ailments encompass a wide spectrum of conditions that influence mood, cognition, and behavior, with depression emerging as one of the most prevalent global challenges in both developed and developing nations. The COVID-19 pandemic has further exacerbated this crisis, leading to widespread emotional distress, job loss, and uncertainty about the future. As a result, depression has surged across all age groups, demanding scalable, empathetic, and data-driven interventions. Artificial Intelligence (AI) through machine learning, natural language processing (NLP), large language models (LLM), deep learning, and Generative AI offers promising avenues for detecting, monitoring, and mitigating the effects of mental stress and depression. Coupled with computer vision, deep learning models can infer emotional states from facial expressions, postures, and behavioral cues, while machine learning algorithms can leverage behavioral histories to generate adaptive, supportive interventions. This tutorial presents a framework that combines responsible AI, digital twin technology, and the Internet of Things (IoT) to deliver ethical, interpretable, and human-centric healthcare solutions, particularly focused on mental health.

Tutorial Overview

This tutorial explores the convergence of AI, IoT, and responsible design in healthcare, providing participants with technical foundations, applied methods, and real-world insights. It is structured into three core modules:

Module 1: Ubiquitous Healthcare via AI-Powered Internet of Things

Participants will explore how ubiquitous IoT systems comprising wearables, smart home devices, and ambient sensors enable continuous, location-independent patient monitoring. This module covers architectural design, data integration strategies, and interoperability standards essential for building scalable and secure health monitoring infrastructures.

Module 2: Predictive Models Utilizing Artificial Intelligence in Healthcare

This module addresses how physiological and behavioral signals collected from IoT devices can be transformed into actionable insights using AI. Topics include predictive modeling for physical and mental health, explainable AI (XAI) for transparency, and fairness-aware techniques to ensure inclusive healthcare. Practical examples, including emotion recognition and behavioral pattern analysis, will be demonstrated.

Module 3: Responsible Artificial Intelligence in Mental Healthcare Systems

Focusing on mental health, this module presents responsible AI methodologies that support the monitoring and management of conditions like depression and anxiety. Discussions will cover bias mitigation, privacy preservation, human-in-the-loop models, and ethical design principles. Real-world applications such as keystroke biomarker systems for Parkinson’s and mood prediction engines will be examined to illustrate responsible AI in action.

Audience Outcome:

By the end of the tutorial, participants will:

• Understand how to design ethical, interpretable, and scalable health monitoring systems using AI and IoT.
• Gain insights into the use of AI for emotion and behavior tracking to address depression and stress.
• Explore responsible AI frameworks for safeguarding mental health data and supporting vulnerable populations.
• Engage with interactive case studies and technical demonstrations to apply concepts in real-world settings.
• Foster cross-disciplinary collaboration for advancing equitable and human-centric digital healthcare solutions.