14 March 2024


Cardiovascular diseases, one of the most common causes of death in the EU, include atrial fibrillation, the most common form of cardiac arrhythmia, which affects over 37 million people worldwide. Atrial fibrillation is also more common in cancer patients and can lead to stroke or heart failure if not treated successfully. The underlying mechanisms of this connection are still unclear. However, virtual twins could provide an ideal framework for their investigation. With the help of digital twin technology, researchers can more accurately predict a patient’s response to treatment and plan customised interventions if necessary.

EDITH partners QMUL and EPFL offer a pathway to personalised in silico treatment planning and model predictions in clinical timeframes. To overcome the challenges of large-scale heart modelling, they are developing a robust open-source pipeline for constructing atrial models based on their previously developed method incorporating atrial regions and fibres. These models were constructed as either two coupled surfaces (a bilayer model) or a volumetric model.

In addition, a cohort of 1000 bi-atrial bilayer and volume models derived from computed tomography (CT) data, as well as models derived from magnetic resonance imaging (MRI) and electro-anatomical mapping data were created. Fibrillation dynamics diverged between the bilayer and volumetric simulations in the CT cohort. Structural and electrophysiological changes in the cardiac tissue due to scar tissue were found to stabilise the re-entries and reduce the impact of the model type (bilayer or volumetric). These models can be used to compare responses to different treatment approaches.

Overall, QMUL and EPFL are successfully testing the integration of the respective resources and cardiac modelling workflows as a proof-of-concept for the creation of virtual human twins (VHTs) and the development of a user-friendly open-source pipeline to generate atrial models from imaging or electro-anatomical mapping data, enabling large-scale in silico clinical studies.