Introduction: Catheter ablation guided by electroanatomical mapping is the cornerstone treatment for complex arrhythmias, yet the software underpinning its analysis remains largely proprietary. This limits reproducibility, restricts methodological transparency, and slows the translation of novel algorithms into clinical and research practice. Although a range of analytical methods exists, including phase mapping and graph-based methods, open-source implementations are scarce, fragmented, and rarely interoperable. We present openDGM, an open-source Python framework that unifies these approaches into a single modular framework for automated arrhythmia source detection from electroanatomical mapping data.
Methods: openDGM is built around six interoperable modules: a Parser supporting clinical systems (CARTO, Rhythmia) and simulation platforms (MonoAlg3D, OpenCARP, OpenEP); a Signal module for filtering, annotation, and local activation time extraction; a Graph module for activation network construction; a Topo module providing topology-based representations; an Algorithm module implementing phase singularity detection, reentry cycle identification, focal source localization, and Helmholtz-based boundary analysis; and a Visualization module based on PyVista that supports interactive data exploration. This modular architecture supports flexible pipeline design and straightforward integration of new methods.
Results: openDGM was validated on simulated, clinical, and experimental datasets, consistently identifying reentrant and focal arrhythmia mechanisms across heterogeneous data sources. Integrating graph-theoretic and topological analyses improved robustness, revealing complex activation patterns not captured by single-method approaches.
Conclusion: openDGM offers a transparent and extensible alternative to proprietary systems. By standardizing electroanatomical data processing and enabling automated mechanism detection, it accelerates algorithm development and promotes collaborative, reproducible research in computational cardiology. The framework is publicly available at https://gitlab.com/opendgm/opendgm.git.