Section: Partnerships and Cooperations

Regional Initiatives

PhD thesis O. Bouhamama

L. Weynans and L. Bear (Liryc) obtained funding for a PhD thesis from the Université de Bordeaux, for a project titled “Méthodes numériques pour la résolution du problème inverse en électrocardiographie dans le cas d’anomalies structurelles du tissu cardiaque.” Candidate O. Bouhamama started in October 2018.


The project “Cardiac Arrhythmia Localization Methods,” granted by the Région Nouvelle-Aquitaine, with matching from funds held by our clinical collaboraters H. Cochet and P. Jaïs, has started. The purpose of this project is to develop a tool that can predict the exit site of an arrhythmia with moderate accuracy (1 cm) in an absolute sense, with respect to the anatomy of the heart in situ, and with a resolution of about 2 mm in a relative sense, with respect to a nearby pacing site. This tool must fulfill the following criteria:

  • it uses only data that are already recorded in the cathlab by other systems: ECG data and electroanatomical mapping data;

  • it must work in nearly real-time; catheter displacement advice must be available within 5 seconds after a paced beat;

  • it must work automatically, requiring the operator only to indicate which ECG data correspond to the target arrhythmia; and

  • it must be safe and easy to operate.

We will in the first place test a number of proposed methods using synthetic data, produced with our realistic models of cardiac electrophysiology and accurate geometric models of different patients. This in-silico testing phase will answer a number of important practical questions. Subsequently we will use offline clinical data, and within 2 years we aim to build a clinical prototype that can be tested (without interfering in the procedure) in the cathlab. In order to work real-time we will initially use very simple methods. However, the clinical prototype and the collectoin of synthetic data that we created will later serve also as a platform to test also more sophisticated inverse methods.