A new national support centre will help researchers use artificial intelligence to understand why people develop multiple long-term health conditions at once, and how to treat them more effectively. The problem is that health data is often locked away in separate systems that cannot talk to each other, and many AI researchers lack the secure computing tools needed to work with sensitive patient information safely. Without a coordinated infrastructure, studies on multimorbidity—where someone has two or more chronic illnesses like diabetes, heart disease, and depression simultaneously—remain fragmented and difficult to verify. The Research Support Facility will provide five things: secure cloud computing certified for health data, data engineering to link and clean messy records from different sources, training and community events including two international conferences and two hackathons, a network connecting patient and public involvement teams across the country, and open-source protocols so that even when the data itself cannot be shared, the methods can be independently checked and reused. If successful, this infrastructure will make AI-driven multimorbidity research faster, more reproducible, and more trustworthy—improving a system that quietly shapes how the NHS diagnoses and treats people with complex health needs.
View original technical description
This proposal brings together the expertise and experience from two national institutes, The Alan Turing Institute and Health Data Research UK (HDRUK), to establish a research support facility (RSF) for the NIHR AI for Multiple Long Term Conditions (Multimorbidity; MLTC-M) programme. It benefits from a combination of expertise in data science methods and health data research; achieved through a collaboration between two of Turing s strategic research programmes, Tools Practices & Systems and Health & Medical Sciences, led respectively by Dr Kirstie Whitaker and Professor Chris Holmes. The RSF will use its convening power and expertise to maximise the scientific impact and potential for patient benefit. This will be focussed around five themes: Robust, secure, reproducible infrastructure. Cloud services are often not certified for sensitive health data and many trusted research environments do not support version control tools for data, software and computational environments. With expertise from the Secure eResearch Platform (SeRP) team we will enable AIM researchers to deliver analyses that can be independently verified and extended. Accessible research-ready data. The RSF team have specialised data engineering expertise to link data from multiple sources, improve data quality and deliver validated, interoperable and reusable data sets to enhance the impact of the MLTC-M research community. Scientific community building and training. We will bring consortia researchers together and spark the interest and excitement of the wider AI community. We will host regular seminars, two international conferences and two collaborative hackathons to inspire data scientists to explore exciting and challenging problems within AI for MLTC-M. Public and patient involvement and engagement (PPI/E). We will build a network connecting PPI/E teams across the AIM Research Collaborations, and stimulate a wider discussion on the opportunities of AI in health research and how an engaged public/patient population can improve the science, AI algorithms, and UK well-being. Sustainability and legacy. We will open source code and protocols, even for sensitive data that can not be made public, following "The Turing Way", a handbook written to make reproducible research "too easy not to do" (https://the-turing-way.netlify.com). We will ensure the investment provides exceptional value for money and longevity in its impact. To achieve this the RSF will run under three overlapping phases Co-design (months 1-3): Working with funded researchers to build a shared vision of the RSF including ways of working and key deliverables. Delivery (months 4-36): Providing community, data, infrastructure and training support for robust, reproducible science with safe and ethical AI led by a vibrant cross-consortia PPI/E team. Legacy (months 4-40): Working from the start to ensure the long term impact of the funding and provide a sustainable foundation for a growing community of MLTC-M researchers. Each theme will be led by a designated Theme Leader, a national expert in their area of data science, together with an early career researcher co-lead. Themes will organise the structure of work and activities in partnership with the funded Research Collaborations to deliver research that is "greater than the sum of its parts" and embeds open, inclusive, ethical values throughout.
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