Our interdisciplinary team brings together expertise in signal processing, machine learning (ML), and clinical cardiology to gain new insights from clinical data with a focus on cardiac time series data. We develop and validate methods for both non-invasive 12-lead electrocardiograms (ECGs) and intracardiac electrograms (EGMs) recorded during ablation procedures, with an eye toward explainable algorithms that can be translated directly into clinical decision support.

Large parts of our work are devoted to rapid, point-of-care risk assessment in the emergency department. We have developed and externally validated a deep learning model that identifies patients in need of coronary revascularization or having an acute type 1 myocardial infarction. Ongoing multicenter validation studies will further establish its generalizability. Parallel projects are focused on predicting disease severity and in-hospital mortality, integrating ECG signals with routinely collected triage data.

Additionally, we systematically analyze EGMs from three-dimensional electroanatomic mapping studies of scar-related cardiac arrhythmias. With our suite of signal-processing and ML algorithms, we aim to identify potential targets for catheter ablation, with the goal of improving precision and outcomes in complex arrhythmia treatment. 

Head of the research group
Dr. Antonius Büscher