M.Sc. Jule Bender

Motivation
Cardiac fibrosis plays a central role in the initiation and maintenance of atrial fibrillation (AF). To conduct in silico studies on AF for treatment planning and risk stratification, synthetic fibrosis models are used. These models need to be realistic and capture relevant variability while being easy, fast and extendible to different fibrosis volumes, and different atrial models.  Therefore, fibrosis modeling encompasses two aspects: the spatial distribution across the atria (macroscopic distribution), and the structure (microscopic distribution). Traditionally, artificial spatial distributions are derived from previous clinical studies or purely artificial structures and combined with random microstructure modeling. Recently, diffusion-model–based generation of macroscopic fibrosis distributions and Perlin-noise–based modeling of fibrosis microstructure was proposed.

Project Description
This project aims to develop a fibrosis model that can be used in virtual cohorts by combining the macroscopic spatial distribution in the atria with recently proposed microscopic structure modeling to generate more physiologically realistic and morphologically diverse AF activation patterns. 

Your tasks:

• Develop a macroscopic fibrosis model based on literature using clinical data.
• Combine the developed model with the recently proposed microscopic modeling approach.
• Conduct electrophysiological simulations to evaluate the developed models in comparison with established modeling approaches in terms of physiologically realistic and morphologically diverse AF activation patterns (atrial activation patterns and ECG)

Notes

• Programming skills in Python are beneficial
• Basic knowledge about cardiac electrophysiology is beneficial
• The thesis can be written in German or English

Weighting can be adjusted according to your interests.