Computer Vision News - January 2022

60 Congrats Doctor Atrial fibrillation (AF) refers to the most common clinical arrhythmia, where the atria activate rapidly and irregularly. Standard catheter ablation protocols to treat AF, however, show sub-optimal long-term AF freedom maintenance, substantiating the need for personalized ablation. Non-invasive mapping techniques provide essential pre- operative planning tools for personalized ablation, which include 12-lead and 252-lead electrocardiograms (ECGs). In my thesis, I developed three algorithms to extract the prognostic values from patient ECGs by using machine learning techniques. Computer models as “digital twins” to aid the development of machine learning algorithms There were several challenges associated with predicting post-ablation outcomes from pre-operative ECGs: there were a limited number of patients as training examples, Yingjing Feng recently completed her PhD in Applied Mathematics at the University of Bordeaux and the IHU-Liryc institute in France. Her expertise lies in bridging machine learning, computational modeling and electrophysiology for atrial fibrillation. Previously, she obtained an MSc in Computing (Machine Learning) with Distinction from Imperial College London. After her PhD, she will start as a research fellow at the University of Birmingham, working on machine learning algorithms for epilepsy modeling. Congrats, Doctor Yingjing!