PhD Studentship: Unlocking the Power of UK Hospital Data: Leveraging Machine Learning and Natural Language Processing for High-Fidelity Clinical Research

Project description:

NHS electronic health records from GP practices have provided a world-leading resource for research for over 20 years. UK hospitals are now also adopting fully electronic patient records, collecting high-fidelity data on patient journeys in hospital and as outpatients for the first time.

This data science focused PhD offers an exciting opportunity to harness these new hospital records for research into complex and serious medical conditions. Prior to being usable for medical research, it is necessary to develop new techniques for correct identification of clinical phenotypes from these complex data. The studentship will provide training on cutting-edge data science techniques with application to real-world clinical data. Core aims include: (a) Applying and combining AI and natural language processing approaches to optimise identification of patient conditions and outcomes: (b) Developing reproducible open-source data processing pipelines to transform raw hospital EHR data into deep phenotyped research-ready datasets; (c) Using diabetes and related conditions to test the accuracy of the approaches and conducting early clinical research using the data.

Analysis will be undertaken within the infrastructure of the Royal Devon University Healthcare NHS Foundation Trust, one of the first five trusts in England to adopt fully electronic records.

Development opportunities will be shaped to meet the student’s needs and interests, including training in data science and epidemiological methods. The student will join the world-leading Exeter diabetes clinical research team and link with the Institute for Data Science and Artificial Intelligence, providing a milieu of research excellence and expertise in addition to the project supervisors’ support. Patients and the public will be engaged to inform the direction of research throughout. The PhD would suit a person with a strong quantitative background, and interests in health data science and clinical research.

Advert information