Development of novel computational approaches for human cardiac mapping data - Improving understanding and treatment of atrial fibrillation

Cardiac arrhythmias are a common pathway for multiple cardiovascular diseases, the leading cause of death in the world.

Ablation of cardiac arrhythmias is a recommended and common treatment option for arrhythmia patients, but success rates are poor. Cardiac electroanatomical mapping systems (EAMs) are used to identify ablation targets.

EAMs generate a wealth of raw data that is ideal for developing algorithms to tailor treatments to an individual patients’ pathophysiology and improve treatment options. However, no platform exists for developing and testing algorithms.

Specific groups have developed closed-source software to identify ablation targets however, these results have not been widely reproducible.

There is an acute need for an open-source and cross-vendor EAM analysis software to facilitate sharing methods, increase rigor and transparency and increase access to advanced processing tools. I will develop expandable open-source software for analysis of EAM data. I will incorporate established and novel analysis methodologies, including machine learning approaches.

Algorithms will be robustly validated using clinical, pre-clinical and in silico datasets.

The software I will develop will enable myself and others to utilise routinely collected clinical data, to better treat cardiac arrhythmia. I will apply these approaches to identify mechanistic drives of atrial fibrillation as an exemplar application.