In-silico trials for patient-specific selection of atrial fibrillation treatment

PI: Rodriguez, Blanca

Department: Computer Science

Atrial fibrillation (AF) is the most common arrhythmia, affecting 1 in 45 people in the UK. After initial treatment, 60% of patients will newly present in the hospital with recurrent AF.

In his DPhil, Albert Dasi has a developed framework for in silico clinical trials to identify optimal therapies for AF. In silico trials refer to the use of human-based modelling and simulation to virtually test medical treatments. Widely accepted in engineering, it is now a promising approach in medicine. The framework focused on the technical challenges involved in (1) constructing large populations of virtual patients that present variability in important clinical features; (2) conducting in-silico trials to evaluate state-of-the-art therapies for AF; (3) developing a decision algorithm that, based on virtual patient features, selects an optimal treatment.

Building on the digital evidence obtained from the simulations, we now aim to translate the developed research tool to clinical practice. For this, Albert will work directly with clinicians in two hospital centres that routinely deliver existing and novel treatments for AF. The immediate step in this direction is the definition of the requirements for impact in clinical practice, including validation of the theoretical findings with clinical data and defining an improved roadmap for risk stratification and therapy selection. With the technical challenges already solved, deep engagement with clinicians will provide the expertise and datasets required for the adaptation and validation of the decision algorithm to boost the efficacy of AF treatment. Similarly, this will increase the confidence in in-silico trials, supporting their integration into clinical practice. This could impact future guidelines of AF treatment, highlighting modelling and simulation as an assisting tool for optimal treatment selection. Great methodological impact is expected from this project, since the framework can be applied to every cardiac disease.