Multiphysics and multiscale modelling, data–model fusion and integration of organ physiology in the clinic: ventricular cardiac mechanics

Radomí­r Chabiniok, Vicky Y. Wang, Myrianthi Hadjicharalambous, Liya Asner, Jack Lee, Maxime Sermesant, Ellen Kuhl, Alistair A. Young, Philippe Moireau, Martyn P. Nash, Dominique Chapelle, David NordslettenInterface 2016 · DOI

TL;DR

This review paper explores the exciting frontier of translational cardiac modeling (TCM), which aims to bridge the gap between fundamental research and clinical practice for heart disease. It highlights how integrating multiphysics, multiscale computational models with patient-specific imaging data is paving the way for improved diagnosis, treatment planning, and device evaluation.

What Problem Does This Paper Address?

Heart and cardiovascular diseases remain a major global health challenge. While our understanding of cardiac physiology and pathology has advanced significantly, translating this knowledge into effective clinical care is hindered by the heart's immense complexity. The heart operates across multiple spatial and temporal scales, involving intricate interactions between electrical, mechanical, and fluid dynamics. Computational modeling offers a powerful tool to unravel these complexities, but its clinical utility is limited by the need to personalize models to individual patients and validate them against real-world data. This paper addresses the challenge of bridging this gap by reviewing the state-of-the-art in cardiac modeling, data assimilation, and their integration for translational applications.

Methods

The paper reviews a wide array of computational modeling approaches for cardiac mechanics, including:

  • Multiscale Modeling: Incorporating phenomena from the subcellular level (e.g., crossbridge dynamics) up to the whole organ level.
  • Multiphysics Modeling: Integrating various physical processes such as electromechanics, fluid-structure interaction (FSI), and poromechanics (for coronary perfusion).
  • Constitutive Equations: Developing mathematical descriptions of passive (anisotropic, hyperelastic/viscoelastic) and active (contraction) myocardial tissue behavior.
  • Data Assimilation and Parameter Estimation: Techniques used to fuse experimental and clinical data (primarily medical imaging) with computational models to personalize them and estimate unknown parameters. This includes methods like variational approaches (e.g., 4D-Var) and sequential methods (e.g., Kalman filters).
  • Image Processing: Essential steps for segmenting anatomical structures and extracting motion information from medical images.

Datasets

The paper discusses the use of various medical imaging modalities as data sources for TCM, including:

  • Echocardiography (ECHO): Widely accessible, providing anatomical and functional information.
  • Computed Tomography (CT): Offers fast acquisition and high spatial resolution, useful for anatomical assessment.
  • Magnetic Resonance Imaging (MRI): Provides detailed functional and tissue characteristic information, including cine MRI, tagged MRI, diffusion tensor imaging (DTI), and MR perfusion.

The paper also references the use of synthetic data and data from experimental studies (e.g., explanted hearts, animal models) for model development and validation.

Key Insights and Ideas

  • The Heart as a Multiphysics System: Understanding cardiac function requires integrating electrical activation, muscle contraction, blood flow, and coronary perfusion, all interacting across different scales.
  • The Power of Personalization: Patient-specific models, derived from medical imaging, are crucial for accurate prediction of individual cardiac behavior and treatment response.
  • Data-Model Fusion is Key: Simply having complex models or abundant data is insufficient; effective integration through data assimilation and parameter estimation is essential for clinical translation.
  • Bridging the Gap is a Collaborative Effort: Successful TCM requires close collaboration between clinicians, engineers, and computer scientists to ensure models address relevant clinical questions and produce clinically actionable outcomes.
  • Uncertainty Quantification is Critical: For models to be trusted in clinical decision-making, quantifying the uncertainty in their predictions due to data variability and model assumptions is paramount.

Key Results

The paper doesn't present specific quantitative results from a single study but rather reviews the collective progress in the field. It highlights that:

  • Computational models have successfully reproduced various aspects of cardiac mechanics, including motion, strain, and pressure-volume relationships.
  • Data assimilation techniques have been demonstrated to personalize cardiac models, leading to improved predictions of cardiac function and therapy response.
  • TCM has shown promise in applications such as device assessment (e.g., heart valves, ventricular assist devices) and therapy planning (e.g., Cardiac Resynchronization Therapy).
  • The development of novel biomarkers derived from model-based analysis is an emerging area with significant potential for diagnosis and prognosis.

Limitations and Future Work

  • Model Complexity vs. Clinical Feasibility: Balancing the need for detailed, multiphysics models with the computational demands and time constraints of clinical practice remains a significant challenge.
  • Data Availability and Quality: While imaging techniques have advanced, limitations in spatio-temporal resolution, reproducibility, and standardization can impact model accuracy.
  • Parameter Identifiability and Uncertainty: Ensuring that model parameters are uniquely identifiable from available data and quantifying the uncertainty in these parameters and model predictions is an ongoing area of research.
  • Validation and Clinical Translation: Rigorous validation of TCM against clinical standards and demonstrating clear clinical benefit over current practices are crucial for widespread adoption.
  • Future Work: The authors emphasize the need for continued development of robust and efficient data assimilation techniques, improved imaging protocols tailored for modeling, and the creation of large, standardized databases for model training and validation. Further research into multiscale modeling, growth and remodeling, and uncertainty quantification is also highlighted as essential for realizing the full potential of TCM.

Summary updated

Abstract

With heart and cardiovascular diseases continually challenging healthcare systems worldwide, translating basic research on cardiac (patho)physiology into clinical care is essential. Exacerbating this already extensive challenge is the complexity of the heart, relying on its hierarchical structure and function to maintain cardiovascular flow. Computational modelling has been proposed and actively pursued as a tool for accelerating research and translation. Allowing exploration of the relationships between physics, multiscale mechanisms and function, computational modelling provides a platform for improving our understanding of the heart. Further integration of experimental and clinical data through data assimilation and parameter estimation techniques is bringing computational models closer to use in routine clinical practice. This article reviews developments in computational cardiac modelling and how their integration with medical imaging data is providing new pathways for translational cardiac modelling.
BibTeX
@inproceedings{chabiniok2016multiphysics,
  author = {Radomí­r Chabiniok and Vicky Y. Wang and Myrianthi Hadjicharalambous and Liya Asner and Jack Lee and Maxime Sermesant and Ellen Kuhl and Alistair A. Young and Philippe Moireau and Martyn P. Nash and Dominique Chapelle and David Nordsletten},
  title = {{Multiphysics and multiscale modelling, data–model fusion and integration of organ physiology in the clinic: ventricular cardiac mechanics}},
  booktitle = {Interface},
  year = {2016},
  doi = {10.1098/rsfs.2015.0083},
}
Figure