By Himashri Deka, Research Scholar, Indian Institute of Technology (IIT), Guwahati and India RISE Fellow 2026
Could technology help identify signs of mental health challenges before they become severe? This question now sits at the centre of my research, but it was not where my engineering journey began.
Like many engineering students in India, I initially viewed engineering through the familiar language of coding, circuits, systems, and technological innovation. Mental health seemed to belong elsewhere, usually to hospitals, counsellors, or specialists, and it was rarely discussed openly in academic spaces. Yet around me, I could see how stress, isolation, academic pressure, and emotional distress often remained unnamed until they had already begun to affect a person’s studies, relationships, or daily life.
That perspective began to change during my time at IIT. Through discussions with my guide, Dr. Hanumant Singh Sekhawat, I was introduced to the idea that engineering and artificial intelligence could contribute to meaningful societal challenges. During one of our conversations, he gave me a copy of Psych 101 by Paul Kleinman. The book offers an accessible introduction to the fundamental concepts, theories, and experiments that have shaped psychology. By exploring how psychologists study human behaviour, cognition, emotions, and decision-making, it encouraged me to think beyond purely technical problems and appreciate the scientific understanding of human experience. Reading it broadened my perspective on how scientific inquiry can be applied not only to understanding the natural world, but also to addressing complex human and societal challenges.
As our discussions continued, I began to look more closely at mental health as a problem of both care and access. In India, many people still delay seeking support because of stigma, lack of awareness, limited availability of trained professionals, cost, or uncertainty about where to begin. For students and young adults in particular, early signs of distress can be mistaken for ordinary pressure or hidden out of fear of being judged. This led me back to the question: could technology help identify signs of mental health challenges before they become severe?
Today, my work focuses on developing AI-driven approaches for mental health awareness and early detection. Specifically, I study human behavioral signals, particularly eye movements and speech to identify patterns associated with mental health conditions.
One area of my research explores eye-movement analysis for the early detection of schizophrenia. Eye movements are closely linked to cognitive processes and brain activity. The way individuals focus, shift attention, scan visual information, or track objects can reveal subtle behavioral markers. While these patterns may be difficult for humans to detect consistently, machine learning models can uncover meaningful signals that support early screening.
Another focus of my research is speech-based depression detection. Human speech conveys far more than words alone. Variations in tone, energy, speaking rate, pauses, and vocal expression often reflect emotional and psychological states. Depression can subtly influence these characteristics, and AI techniques can help identify such changes through the analysis of voice recordings.
What fascinates me most is that these indicators emerge from ordinary human behavior. We speak and move our eyes every day without conscious thought, yet these seemingly routine actions can reveal silent struggles that individuals may find difficult to express.
The purpose of this research is not to replace doctors, psychologists, or therapists. Mental healthcare fundamentally depends on human empathy, professional expertise, and meaningful interpersonal support. Rather, I see AI as a complementary tool, one that can assist in early screening, raise awareness, and encourage timely intervention. Early detection is particularly important in mental health. Many individuals postpone seeking help because of stigma, limited awareness, or fear of being judged. By the time support is received, the condition may already have affected their academic performance, career, relationships, or overall quality of life. Technologies capable of identifying early warning signs could help bridge this gap and facilitate access to care at a much earlier stage.
This potential is especially significant in India, where mental health resources remain limited relative to the size of the population and where support is often concentrated in urban or institutional settings. For someone who is unsure whether their distress is serious enough to seek help, or who does not have easy access to a counsellor or psychiatrist, AI-enabled screening tools could offer a first point of awareness. In the future, such tools may help make preliminary assessment more accessible, scalable, and affordable. Working in this field has also transformed my understanding of technology itself.
Research is often measured through algorithms, accuracy metrics, and technical publications. Yet behind every dataset lies a human story. This perspective serves as a constant reminder that technology should not only be intelligent, but also responsible, ethical, and meaningful. My journey has shown me that engineering can create impact far beyond its traditional applications. By combining signal processing, machine learning, and data-driven analysis, I hope to contribute to a future where mental health conditions are detected earlier, understood more deeply, and discussed more openly.
Sometimes the signs are visible in the movement of the eyes. Sometimes they are heard in the nuances of a voice. And sometimes, recognizing those signs early can be the first step toward healing.
Research Scholar, Indian Institute of Technology (IIT), Guwahati and India RISE Fellow 2026
