CDTnet PhD Fellowship F5: Digital Twin Models to Infer Anatomical and Functional Parameters at Scale

Cardiovascular digital twins have the potential to transform medicine by providing personalised models of anatomy and physiology that support diagnosis, prognosis and treatment planning. However, creating these models currently requires complex and time‑complicated workflows that limit their use in large patient populations and routine clinical practice. This project aims to develop new methods for personalising cardiovascular digital twins  doctoral candidate will create computational approaches capable of estimating anatomical, mechanical and electrophysiological parameters from large imaging datasets while addressing the challenges of limited patient‑specific data and uncertainty in model predictions.

A key objective is to identify the appropriate balance between model complexity and clinical applicability, allowing robust digital twins to be generated efficiently for large numbers of individuals. The project will make extensive use of the UK Biobank, one of the world's largest health research resources, with access to imaging and functional data from about 100,000 participants. Using these data, the candidate will develop population‑based reference values for cardiac anatomy and function, including parameters such as contractility, passive stiffness and electrical conduction properties.

The project will also investigate how these parameters change over time by studying longitudinal data from thousands of participants and constructing digital twin trajectories that describe cardiovascular ageing and disease progression. An important aspect of the research will be the study of variability across populations, including differences related to sex and ethnicity. The resulting methodologies and datasets will support many of the other CDTnet projects by providing population‑scale digital twin frameworks that can be applied to heart failure, arrhythmias and valve disease.

Ultimately, the project aims to help make personalised cardiovascular modelling more accessible, scalable and clinically useful.

• Maastricht University, Netherlands (2 months): training in the Circ Adapt cardiovascular modelling platform and large-scale model personalisation.
• University of Zagreb School of Medicine and IDIBAPS, Spain (short stays): exposure to clinical data and workflows related to heart failure, arrhythmias and valve disease.
• University of Zaragoza, Spain (short stay): integration of clinical datasets into digital twin pipelines and support for related CDTnet projects.
• GE Vingmed Ultrasound, Norway (2 months): investigation of digital twin personalisation from echo cardiographic data.

• Bachelor Degree or equivalent, Evidence of ability to work with computational intensive analysis routines
• Evidence of skills in image or bio-signal (e.g. ECG) analysis
• Evidence of skills in model personalization: parameter estimation, uncertainty quantification, model identifiability

• MSCA Mobility Rule:
  You must not have lived or worked in the United Kingdom for more than 12 months in the 3 years before recruitment.
• MSCA Eligibility Rule:
  You must not already hold a doctoral degree and must be eligible to enrol in the PhD programme at King's College London.