Multiphysics modelling and neural networks synergy to enhance the outcome of robotic catheter ablation for the treatment of atrial fibrillation

Atrial fibrillation is the most common cardiac arrhythmia affecting more than 7.6 million people in Europe, while future projections estimate a worldwide increase by over 60%. Catheter ablation has evolved over the last decades to an important therapeutic modality for patients with atrial fibrillation. Despite the fact that modern robotic ablation systems made precise catheter navigation possible, the success rate remains limited and the risk for complications is significant. As a result, atrial fibrillation remains a serious health issue, influencing the quality of life, and the source of very important socioeconomic problems. The vision of PhyNeTouch project (reads: fine-touch) is to capitalize on the recent advancements in computational modelling and artificial intelligence in order to enhance the safety of robotic catheter ablation procedure and improve its success rate. Novel multiphysics modelling techniques based on meshfree methods will be combined for the first time with physics-informed neural networks to advance the state-of-the-art and produce accurate bioheat distribution maps in real-time which currently are not available in the clinical setting. Predicting bioheat distribution in the tissue during the operation will assist the decision making and will allow improving the outcome of robotic catheter ablation. In this way, PhyNeTouch aims to contribute to the healthy ageing of citizens of EU and the rest of the world, as well as to reduce the socioeconomic burden, by reducing the complexity of this procedure and making it accessible to a wider public.