Understanding the clinical needs for adoption of cardiovascular modelling in precision medicine

Digital Twins for healthcare are virtual replicas of systems (e.g. organs) that are continually updated via data streams and hold a bidirectional relationship with their physical twin. This technology has the potential to revolutionise diagnostics, prognostics, and surgery planning. However, its clinical adoption has been slow. Our medium-term ambition is to build a sustained, externally funded programme delivering impactful cardiovascular digital twins that are genuinely deployable in routine care. This sits at the interface of modelling, software and simulation engineering, clinical decision-making, and organisational studies. It is important because digital twins are now technically mature but rarely cross the translational gap: they are poorly aligned with clinical schedules, trained on controlled state-of-the-art data, and insufficiently tailored to how decisions are actually made in clinics and theatres. In contrast, we want to create a design framework in which clinical workflows, timings, information needs and constraints explicitly shape what a digital twin should be. This pilot project is the first step to enable this ambition: it will establish feasibility, help design research methodology, and generate initial empirical data on workflows and user needs in three concrete use cases from routine practice of the congenital heart disease (CHD) team at Evelina Children’s Hospital: (1) surgical planning for complex cardiac repair, (2) catheter planning for atrial septal defect closure and (3) assessment of functional metrics in CHD, e.g. pulmonary vascular resistance. These data will then be used to outline challenges for digital-twin designs that are fit for purpose in real clinical environments.

Therefore, we aim to map and formalise clinical workflows in CHD care to answer the following research questions:

• What are appropriate methods for discovering the barriers and challenges for integrating digital twins into clinical work practice?
• What type of information is expected from the digital twin by a clinical user and how should this be communicated so digital twins are trusted and used?
• At what points in the clinical workflow can cardiovascular digital twins be most impactful?
• What are realistic data and time budgets for these models in different clinical pathways?

Answering these research questions will directly inform the design of usable and impactful cardiovascular digital twins. We will demonstrate proof of concept in use cases from CHD settings as these provide a rich selection of clinical pathways, where digital twins could enhance clinical decision-making substantially.

This research is a collaboration between the following partners: