A computer will scan anonymised NHS mental health records for emotional warning signs hidden in clinical notes, searching for patterns of loneliness or distress that might predict a crisis before it happens. Most research into severe mental illnesses like schizophrenia or bipolar disorder focuses on biology or medication. This project addresses a gap: the emotional and social factors—such as isolation, difficult relationships, or lack of access to community spaces—that shape how people become unwell and recover. These subjective experiences are widely recognised but rarely studied systematically. If successful, the research could change how mental health services detect and respond to early signs of deterioration. Instead of waiting for a crisis, clinicians might receive automated alerts from AI tools that spot emotional distress in patient records. Planners could use data on how local environments—like access to parks or cultural venues—affect outcomes for people with SMI. A third strand tests whether real-time reporting of mood and social contact via smartphones can help predict relapses. People with lived experience co-lead the work, keeping it grounded in real priorities. The ultimate aim is more compassionate, timely care and reduced stigma.
View original technical description
Severe and enduring mental illnesses (SMI) such as schizophrenia, bipolar disorder, severe depres-sion, and complex emotional needs, can deeply affect how people think, feel, and relate to others. While research often focuses on biology or medication, much less attention has been given to the emotional and social factors that shape mental health. Experiences like loneliness, emotional dis-tress, and the stress of navigating difficult social environments can play a major role in how people become unwell, what support they receive, and how they recover. This project, called WISDOM, focuses on understanding these socioemotional drivers of mental illness. I will look at how people’s emotions, relationships, and environments influence the course of SMI and how these insights can improve care and support. This person-centred approach high-lights subjective experiences to which we can all relate, yet are not considered enough in mental health research. The project will have three parts: First, I will use cutting-edge language models (like AI tools) in a trusted research environment to search for emotional and social warning signs in anonymised clin-ical records to help services better detect and respond to distress earlier. Second, I will examine how people’s local environment like social isolation or access to cultural spaces affects outcomes for those with SMI which can guide planning and interventions. Third, I will explore whether col-lecting real-time information on how people feel and who they connect with during daily life can help predict or prevent mental health crises. People with lived experience of mental illness will guide and co-lead the project throughout, en-suring the research stays grounded in real-life priorities. Together, we will develop new knowledge, tools, and public resources that improve mental health services, reduce stigma, and support more compassionate care.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know