The openCARP Simulation Environment for Cardiac Electrophysiology

Gernot Plank, Axel Loewe, Aurel Neic, Christoph M. Augustin, Yung-Lin Huang, Matthias A. F. Gsell, Elias Karabelas, Mark Nothstein, Anton J. Prassl, Jorge Sánchez, Gunnar Seemann, Edward J. VigmondbioRxiv 2021 · DOI

TL;DR

This paper introduces openCARP, an open-source simulation environment designed to standardize and democratize computational cardiac electrophysiology (CEP) modeling. By combining a powerful C++ simulator with a Python-based workflow framework (carputils), openCARP aims to make complex, multi-scale cardiac simulations more accessible, reproducible, and efficient for researchers worldwide.

What Problem Does This Paper Address?

Computational modeling of cardiac electrophysiology is a vital tool for understanding heart function, developing new therapies, and advancing precision cardiology. However, the field has historically lacked a community-standard software platform. This has led to fragmented development, difficulties in sharing and reproducing results, and a steep learning curve for new researchers. Existing commercial software is often limited in speed and capability, while academic in-house codes, though powerful, are difficult to maintain and share. This gap hinders the broader adoption and impact of in silico CEP research.

Methods

The openCARP environment is built around two core components:

  • openCARP simulator: A high-performance C++ simulator that implements state-of-the-art numerical methods (Finite Element Method, FEM) for solving the mono- and bidomain equations that describe cardiac electrical activity. It leverages parallel computing (PETSc, LIMPET) for efficiency and supports various time-stepping schemes.
  • carputils framework: A Python-based framework that provides a standardized way to define, execute, and manage in silico experiments. It automates complex workflows, from parameterization and simulation setup to post-processing and visualization, making it easier to share and reproduce studies.

The system emphasizes:

  • Accessibility: Binary packages, Docker containers, and detailed documentation are provided for easy installation across various platforms, including high-performance computing (HPC) clusters.
  • Interoperability: openCARP can import and export data in standard formats like VTK and CellML, facilitating integration with other tools.
  • Reproducibility: Standardized workflows and data management through carputils are key to ensuring that simulations can be reliably reproduced.
  • Usability: A focus on intuitive interfaces and comprehensive documentation aims to lower the barrier to entry for researchers.

Datasets

The paper does not explicitly list specific datasets used for training or evaluation in the traditional machine learning sense. Instead, it highlights the use of:

  • Anatomical models: The system supports the use of anatomical models derived from medical imaging (e.g., CT, MRI) to create realistic patient-specific or population-based geometries.
  • Cellular models: It integrates a library of commonly used ionic models (e.g., Courtemanche et al., O'Hara et al.) and supports importing models via CellML.
  • Verification benchmarks: openCARP incorporates a suite of verification benchmarks, including the CEP N-version benchmark, to ensure the correctness of its numerical implementations.

Key Insights and Ideas

  • The need for standardization is critical: The paper strongly argues that the lack of a common, open-source platform is a major bottleneck for the CEP modeling community.
  • Bridging the gap between complexity and usability: openCARP aims to provide the power of advanced simulation tools without the prohibitive complexity, making them accessible to a wider range of researchers.
  • Workflow automation is key to reproducibility: The carputils framework is presented as a crucial component for automating and standardizing the entire simulation pipeline, from setup to analysis.
  • Community-driven development is essential for sustainability: The paper emphasizes the importance of a vibrant user and developer community, supported by accessible documentation, training, and communication platforms.
  • Multi-scale modeling is the future: openCARP is designed to handle simulations from the single-cell level up to the whole-body level, reflecting the interconnectedness of cardiac function.

Key Results

The paper focuses on presenting the openCARP environment and its capabilities rather than reporting specific quantitative results from extensive simulations. However, it demonstrates the system's ability to:

  • Automate complex workflows: The carputils framework is shown to streamline multi-scale simulations, from single-cell models to organ-level simulations of atrial fibrillation.
  • Support diverse simulation scenarios: Examples are provided for cellular level analysis (e.g., AP phenotypes, restitution), tissue level investigations (e.g., conduction velocity, heterogeneity), organ level modeling (e.g., biatrial models), and body level simulations (e.g., ECG generation).
  • Facilitate model development and sharing: The open-source nature and standardized workflows are designed to foster collaboration and the reuse of existing models and experiments.

Limitations and Future Work

  • Limitations:
    • While integration testing is extensive, the paper notes that unit testing is not performed at the current stage of development.
  • The authors acknowledge that while openCARP provides many elements for building workflows, a complete, integrated workflow for tasks like anatomical model generation is still under development.
  • The full potential of openCARP for clinical applications is still being explored, and regulatory bodies are only beginning to encourage the use of numerical models.
  • Future Work:
    • Continued development of automated pipelines for anatomical model generation.
  • Further integration of community-contributed models and features.
  • Expansion of documentation and training materials.
  • Development of mechanisms for easier integration with other simulation and analysis tools.
  • Continued efforts to ensure the long-term sustainability and accessibility of the openCARP ecosystem.

Summary updated

Abstract

Abstract Background and Objective Cardiac electrophysiology is a medical specialty with a long and rich tradition of computational modeling. Nevertheless, no community standard for cardiac electrophysiology simulation software has evolved yet. Here, we present the openCARP simulation environment as one solution that could foster the needs of large parts of this community. Methods and Results openCARP and the Python-based carputils framework allow developing and sharing simulation pipelines which automate in silico experiments including all modeling and simulation steps to increase reproducibility and productivity. The continuously expanding openCARP user community is supported by tailored infrastructure. Documentation and training material facilitate access to this complementary research tool for new users. After a brief historic review, this paper summarizes requirements for a high-usability electrophysiology simulator and describes how openCARP fulfills them. We introduce the openCARP modeling workflow in a multi-scale example of atrial fibrillation simulations on single cell, tissue, organ and body level and finally outline future development potential. Conclusion As an open simulator, openCARP can advance the computational cardiac electrophysiology field by making state-of-the-art simulations accessible. In combination with the carputils framework, it offers a tailored software solution for the scientific community and contributes towards increasing use, transparency, standardization and reproducibility of in silico experiments.
BibTeX
@misc{plank2021i,
  author = {Gernot Plank and Axel Loewe and Aurel Neic and Christoph M. Augustin and Yung-Lin Huang and Matthias A. F. Gsell and Elias Karabelas and Mark Nothstein and Anton J. Prassl and Jorge Sánchez and Gunnar Seemann and Edward J. Vigmond},
  title = {{The <i>openCARP</i> Simulation Environment for Cardiac Electrophysiology}},
  year = {2021},
  doi = {10.1101/2021.03.01.433036},
  howpublished = {bioRxiv},
}
Figure