Cardiac diseases are a global health burden and the number one cause of death in Germany. In the case of atrial fibrillation (AF), the treatment is often insufficient because of the lack of knowledge about the underlying disease mechanisms that cause stroke. Atrial cardiomyopathy (AtCM) leads to AF and is suspected to be responsible for the increased risk for stroke instead of AF itself. The investigation of the relationship between AtCM and stroke events could lead to different therapies, reducing the risk for stroke.
Studies by Jadidi et al. showed the potential of the amplified P-wave duration in the ECG as a predictor for AtCM. As a result, the automatic detection and calculation of P-wave features in the ECG utilizing clinical data from the Universitäts-Herzzentrum Bad Krozingen and machine learning algorithms is aimed.

Added to the clinical data, synthetic data will be used to analyze the influence of electrode positions to find the optimal electrode placement for the P-wave. A clinical validation of the AtCM prediction using this optimal electrode placement, will be conducted using the invasive activation times and atrial electroanatomical maps as ground truth.

A further project during this doctorate will be the development and utilization of machine learning algorithms on a large dataset consisting of more than 1,000,000 digital ECGs and other additional patient data recorded by the Universitäts-Herzzentrum Bad Krozingen. The aim is to predict cardiovascular diseases.