������app

Summary

Our research focuses on the following areas:

• Behavioural and preventative health
• Age-related chronic conditions (diabetes, heart disease, dementia) and healthy ageing/longevity
• Primary healthcare and care in the community/at home
• Affordable AI solutions for resource-limited populations and health systems
• Applications of generative AI and large language models

Current research

Digital diagnostics for chronic disease research in India

We have established a long-standing longitudinal cohort study in India, the Andhra Pradesh Children and Parents' Study (APCAPS), which follows approximately 7,000 participants over more than two decades. Originating from a childhood nutrition intervention, APCAPS has evolved into a unique platform for life-course epidemiological research in a rural, socioeconomically transitioning population. The cohort has undergone multiple waves of detailed phenotyping, including assessments of environmental exposures, lifestyle behaviours, and health outcomes.

Our research group is now leveraging this rich dataset to develop and evaluate digital diagnostic tools for chronic disease detection. By employing machine learning techniques to analyse body composition scans, vocal characteristics, and gait patterns, we are exploring novel digital phenotypes for conditions such as diabetes, cognitive decline, and frailty. In parallel, we are integrating wearable sensors, retinal imaging, ECG, and multi-omics profiling to create scalable, low-cost diagnostic solutions suitable for resource-constrained settings.

Learn more about APCAPS study.

LLM-powered technology platform to support management of overweight adults in Thailand

We are co-developing a large language model (LLM)-enabled digital health platform to support the management of overweight and obesity among adults in Thailand. This intervention draws on advances in artificial intelligence to deliver personalised, scalable behavioural support through digital modalities. Core components include an AI-powered chatbot, interactive mobile applications, group session videos, and virtual communication channels with healthcare providers and peer supporters. These features are introduced incrementally over a 12-month period to facilitate sustained user engagement.

Our approach is grounded in behavioural science and informed by close collaboration with Thailand’s Ministry of Public Health, primary care teams, and service users. Particular attention is given to ensuring accessibility for underserved groups, including older adults, those residing in rural or remote areas, and individuals with limited digital literacy. By integrating behavioural coaching, real-time personalised feedback, gamification, and wearable-linked self-monitoring, we aim to foster long-term, health-enhancing behaviours. This initiative is intended to serve not only as a national strategy for obesity reduction but also as a transferable model for equitable, digitally-enabled non-communicable disease management across other low- and middle-income settings.

Mobile AI platform to support management of diabetic foot ulcers in Nepal, India, Thailand, and Indonesia

We are leading a multi-country research initiative to enhance the prevention and management of diabetic foot ulcers (DFU) through mobile artificial intelligence technologies, with a focus on rural areas across Nepal, India, Thailand, and Indonesia. Our programme seeks to strengthen primary healthcare capacity by introducing affordable, scalable tools—including pulse oximeters, mobile-phone-compatible thermographic imaging, AI-driven wound assessment applications, and automated text-based reminders to support patient self-care.

The research is structured into three interlinked work packages. Initially, we are conducting in-depth context analyses, including burden estimation, healthcare provider assessments, and care pathway mapping in each country. This will inform the participatory co-design of context-appropriate technological solutions with input from patients, providers, and policy actors. Subsequently, we will evaluate the feasibility of implementation within routine care and estimate the health and economic impact of wider scale-up. The programme is underpinned by a strong emphasis on equity, sustainability, and capacity building, including fellowships and doctoral training for early-career researchers in the region. Our ultimate aim is to establish a robust, replicable model for AI-supported chronic disease care in low-resource settings.

PhD student projects

• Use of machine learning to:
  • Study multimorbidity patterns in India, Thailand, Japan
  • Diagnose dementia from speech in rural India
  • Optimise diabetes treatment in Thailand
• Use of generative AI to:
  • Study trends in NCDs from social media posts in India
  • Provide culturally tailored behavioural support in LMICs
  • Improve diabetes screening in Indonesia
  • Address ethnic inequalities in healthcare access in the UK
  • Improve stroke care in Malaysia

Other research

We are also conducting a series of reviews to bridge research gaps in the implementation of artificial intelligence across multiple domains, including primary care, radiology, cardiology, diabetes, research methodologies, behaviour change, community health worker engagement, cost-effectiveness, and ethical and governance frameworks.