Research pillar 1

AI & Digital Phenotyping

Using advanced AI to identify patterns, phenotypes and trajectories across complex health data, and to predict outcomes so that action can be earlier and more targeted.

Focus areas

Machine learning and foundation models
Digital phenotyping
Disease trajectory modelling
Clinical prediction and risk stratification
Explainable, robust and trustworthy AI

Projects & grants: Details to be confirmed before publication. See also Funded research.

Videos

Related films

How artificial intelligence (AI) can help people with health conditions — Plain English Summary

Selected publications

Evidence for this pillar

Large language models for thematic analysis in healthcare research: A blinded mixed-methods comparison with human analysts

Hill C, Dahil A, Simpson G, Hardisty D, Keast J, Kumar Pinn C, Dambha-Miller H. PLOS Digital Health, 2026.

View DOI

Artificial intelligence-driven exercise programmes in personalising the management of multimorbidity

Keast J, Simpson G, Smith L, Dambha-Miller H. BJGP Open, 2025.

View DOI

Equity considerations in AI-enabled tools for healthcare (group output)

See Impact page and News for further AI-related outputs as DOIs are confirmed.

View DOI