Jagir R. Hussan

Using CellML with OpenCMISS to Simulate Multi-Scale Physiology.

OpenCMISS is an open-source modeling environment aimed, in particular, at the solution of bioengineering problems. OpenCMISS consists of two main parts: a computational library (OpenCMISS-Iron) and a field manipulation and visualization library (OpenCMISS-Zinc). OpenCMISS is designed for the solution of coupled multi-scale, multi-physics problems in a general-purpose parallel environment. CellML is an XML format designed to encode biophysically based systems of ordinary differential equations and both linear and non-linear algebraic equations. A primary design goal of CellML is to allow mathematical models to be encoded in a modular and reusable format to aid reproducibility and interoperability of modeling studies. In OpenCMISS, we make use of CellML models to enable users to configure various aspects of their multi-scale physiological models. This avoids the need for users to be familiar with the OpenCMISS internal code in order to perform customized computational experiments. Examples of this are: cellular electrophysiology models embedded in tissue electrical propagation models; material constitutive relationships for mechanical growth and deformation simulations; time-varying boundary conditions for various problem domains; and fluid constitutive relationships and lumped-parameter models. In this paper, we provide implementation details describing how CellML models are integrated into multi-scale physiological models in OpenCMISS. The external interface OpenCMISS presents to users is also described, including specific examples exemplifying the extensibility and usability these tools provide the physiological modeling and simulation community. We conclude with some thoughts on future extension of OpenCMISS to make use of other community developed information standards, such as FieldML, SED-ML, and BioSignalML. Plans for the integration of accelerator code (graphical processing unit and field programmable gate array) generated from CellML models is also discussed.

Open Access Integrated Therapeutic and Diagnostic Platforms for Personalized Cardiovascular Medicine

It is undeniable that the increasing costs in healthcare are a concern. Although technological advancements have been made in healthcare systems, the return on investment made by governments and payers has been poor. The current model of care is unsustainable and is due for an upgrade. In developed nations, a law of diminishing returns has been noted in population health standards, whilst in the developing world, westernized chronic illnesses, such as diabetes and cardiovascular disease have become emerging problems. The reasons for these trends are complex, multifactorial and not easily reversed. Personalized medicine has the potential to have a significant impact on these issues, but for it to be truly successful, interdisciplinary mass collaboration is required. We propose here a vision for open-access advanced analytics for personalized cardiac diagnostics using imaging, electrocardiography and genomics.

A mean-field model of ventricular muscle tissue.

A theoretical model of the cross-linking topology of ventricular muscle tissue is developed. Using parameter estimation the terms of the theoretical model are estimated for normal and pathological conditions. The model represents the anisotropic structure of the tissue, reproduces published experimental data and characterizes the role of different tissue components in the observed macroscopic behavior. Changes in the material parameters are consistent with expected structural changes and the model is extended to reproduce force-Calcium relationships. Model results are invoked to argue that semisoft behavior and the material axis anisotropy arise from the constraints on the extracellular matrix cross-linking topology.

ICMA: an integrated cardiac modeling and analysis platform

Summary: ICMA, a software framework to create 3D finite element models of the left ventricle from cardiac ultrasound or magnetic resonance imaging (MRI) data, has been made available as an open-source code. The framework is hardware vendor independent and uses speckle tracking (endocardial border detection) on ultrasound (MRI) imaging data in the form of DICOM. Standard American Heart Association segment-based strain analysis can be performed using a browser-based interface. The speckle tracking, border detection and model fitting methods are implemented in C++ using open-source tools. They are wrapped as web services and orchestrated via a JBOSS-based application server. Availability and implementation: The source code for ICMA is freely available under MPL 1.1 or GPL 2.0 or LGPL 2.1 license at https://github.com/ABI-Software-Laboratory/ICMA and a standalone virtual machine at http://goo.gl/M4lJKH for download. Contact: r.jagir@auckland.ac.nz Supplementary information: Supplementary materials are available at Bioinformatics online.

Data-driven reduction of a cardiac myofilament model

This manuscript presents a novel, data-driven approach to reduce a detailed cellular model of cardiac myofilament (MF) for efficient and accurate cellular simulations towards cell-to-organ computation. Based on 700 different sarcomere dynamics calculated using Rice model, we show through manifold learning that sarcomere force (SF) dynamics lays surprisingly in a linear manifold despite the non-linear equations of the MF model. Then, we learn a multivariate adaptive regression spline (MARS) model to predict SF from the Rice model parameters and sarcomere length dynamics. Evaluation on 300 testing data showed a prediction error of less than 0.4 nN/mm2 in terms of maximum force amplitude and 0.87 ms in terms of time to force peak, which is comparable to the differences observed with experimental data. Moreover, MARS provided insights on the driving parameters of the model, mainly MF geometry and cell mechanical passive properties. Thus, our approach may not only constitute a fast and accurate alternative to the original Rice model but also provide insights on parameter sensitivity.

Inferring intra-cellular mechanics using geometric metamorphosis: A preliminary study

Mechanotransduction plays an important role in sub-cellular processes and is an active area of research. Determining the forces/strains that the intra-cellular structures experience is vital for developing quantitative models of cellular behavior. Established techniques such as traction force microscopy, digital image correlation etc. track surface forces and kinematics of intra-cellular structures. However, difficulties arise when cells cannot be seeded on micro-patterned substrates or the intra-cellular structures vary (unstable landmarks). Here, we applied geometric metamorphosis, a global image registration method, to determine the kinematic profile of a cell during cell division. The method does not require stable landmarks, the registration is non-local in nature and constraints such as volume conservation can be enforced. The cell wall was tracked over time and a sequence of transformations relating the cell wall at the start of cytokinesis to the configuration prior to the daughters completely separate was determined. These transformations are associated with a scalar metric and a statistical atlas describing the wall kinematics from multiple trackings of the wall shape is constructed. Using these transformations, the cellular kinematics can be described using a Lagrangian frame of reference and the evolution of a material point property can be easily modeled. To demonstrate this, we use the kinematic data derived from the atlas along with a model of stress-fiber (de)formation dynamics to simulate the stress-fiber configuration as the cell domain deforms.

Efficient solutions of cardiac membrane models using novel unsupervised clustering algorithm

We present a method to efficiently solve cardiac membrane models using a novel unsupervised clustering algorithm. The unsupervised clustering algorithm was designed to handle repeated clustering of multidimensional objects with rapidly changing properties. A Modified Trie datastructure that allowed efficient search, scalable and distributed assembly of the result was designed. The method was applied to solve monodomain models of cardiac tissue with highly non-linear reaction elements. We demonstrate the versatility and advantages of using the method by subjecting the tissue to various spatial excitation patterns.