Clinical need: Heart failure with preserved ejection fraction (HFpEF) is a condition where the heart does not meet the needs of the body because it is not properly filling during diastole. It requires numerous visits through these healthcare units, and it thus generates a wealth of data that ends up scattered in different information systems. There is a need for a better management of this condition by focusing on the patient journey.
Novelty: This project envisions the ability to early detect, diagnose and monitor HFpEF through the personalization and continuous update of the digital twin of each patient. In healthcare the digital twin refers to a comprehensive and coherent integration of the clinical data acquired over time for an individual using mechanistic and statistical models. The strength of this approach is thus the ability to combine the strengths of AI discovery (inductive reasoning) and mechanistic inference (deductive reasoning) to characterise disease trajectories.
Aims and objectives: The overarching aim is to build a digital twin framework that describes trajectories in HFpEF. The specific objectives are (1) To define the specifications of a model to define HF trajectories; (2) To describe the baseline healthy characteristics of such model; (3) To describe trajectories into HFpEF; (4) To enrich such trajectories with advanced model-derived biomarkers of diastology; and (5) To test the ability of such model and trajectories to predict disease evolution and major adverse cardiovascular events.