NHS England Data Science PhD Internships

Building towards a Primary Care Digital Twin

Keywords: Synthetic, Digital Twins, Tabular

Need: A digital twin (DT) is a digital representation of an intended or actual real-world physical product, system, or process that serves as the effectively indistinguishable digital counterpart of it for practical purposes. Recently, DTs in healthcare have seen increased usage for modelling physical components (e.g. hearts) or for personalised healthcare (e.g. creating an individual twin and investigating the impact of a treatment).

There is a need to understand how models and algorithms will be able to interact with an individual patient record in a safe and ethical manner. This project would look to explore the components required to create a digital twin of a patient through the lens of their associated data record and the safety harnesses required to interact these data with prognosis and prediction models.

An extension of this work would be to create this digital twin of the patient record across multiple data collection systems to reflect the realities around how the data are recorded and stored in the NHS. This would require simulating this patient interacting with a sub-set of the healthcare system over time and recording their activity and medical updates in a variety of systems and formats (e.g. physical patient letters, scans on a diagnostic system, activity on a patient administration system, patient record as a FHIR bundle in a primary care system, etc..).

Current Knowledge/Examples & Possible Techniques/Approaches:

Related Previous Internship Projects: Our extended SynPath work is being developed to simulate patients following a trajectory through a system and generating associated data.

Enables Future Work: This work would feed into an ongoing project to demonstrate primary care innovations in a safe and transparent way.

Outcome/Learning Objectives: Report and open codebase demonstrating the deployment of predictive models on the digital twin of patient record.

Datasets: Open data for non-disclosive patient demographics. Possibility of using OpenSafely for system characterisation.

Desired skill set: When applying please highlight any experience around work with synthetic data, primary care, python coding experience and software development (including any coding in the open), and any other data science experience you feel relevant.

Return to list of all available projects.