Contents

📒 Plank 2008

From mitochondrial ion channels to arrhythmias in the heart: computational techniques to bridge the spatio-temporal scales1

Sciwheel

Introduction: simplifications of CMC models

  • simple geometry (ellipsoid-shaped)
  • under-representation
  • adjusting tissue conductivity tensors. Coarse granularity => conduction block
  • homogeneous continuum in the myocardium
  • ion transport kinetics are modelled in a simplified fashion, phenomenologically represented parameters do not directly correspond to actual molecular structures or processes

Review of tissue- and organ-level modelling techniques

The bidomain equations

  • homogenization of cardiac tissue, replacing discrete components of the intracellular and extracellular tissue matrix, such as cells and gap junctions, with a continuum. expit are conductivity tensors. Conduction velocity is faster along the direction of the fibre orientation, and slower in a direction transverse to it.

$$ \nabla \cdot \overline{\sigma}_{ \mathrm{i} } \nabla \phi_{ \mathrm{i} }=\beta I_{\mathrm{m} } $$ $$ \nabla \cdot \overline{\sigma}_{ \mathrm{e} } \nabla \phi_{ \mathrm{e} }=-\beta I_{ \mathrm{m} } $$ $$ I_{\mathrm{m} }=C_{\mathrm{m} } \frac{\partial V_{\mathrm{m}} }{\partial t}+I_{\mathrm{ion} }\left(V_{\mathrm{m} }, v\right)-I_{\mathrm{trans} } $$ $$ V_{\mathrm{m}}=\phi_{\mathrm{i}}-\phi_{\mathrm{e}} $$ $$ \nabla \cdot\left(\overline{\sigma}_{\mathrm{i}}+\overline{\sigma}_{\mathrm{e}}\right) \nabla \phi_{\mathrm{e}}=-\nabla \cdot \overline{\sigma}_{\mathrm{i}} \nabla V_{\mathrm{m}}-I_{\mathrm{e}} $$ $$ \nabla \cdot \overline{\sigma}_{\mathrm{i}} \nabla V_{\mathrm{m}}=-\nabla \cdot \overline{\sigma}_{\mathrm{i}} \nabla \phi_{\mathrm{e}}+\beta I_{\mathrm{m}} $$ A Poisson problem has to be additionally solved for the medium $$ \nabla \cdot \sigma_{\mathrm{b}} \nabla \phi_{\mathrm{e}}=I_{\mathrm{e}} $$

  • Boundary conditions: Typically, grounding electrodes are used in the extracellular domain, i.e. nodes in the mesh are chosen where ϕe is fixed at zero value

Discretization schemes and issues

  • FDM, FVM, FEM
  • spatial extent of the wavefront is in the range of 0.2–0.7 mm
  • Courant–Friedrichs–Lewy (CFL) condition for time step limit for numerical stability in FE method $$ \Delta t \leq \frac{\beta C_{\mathrm{m}} \Delta x^{2}}{2\left(\sigma_{l}+\sigma_{t}\right)} $$
  • More efficient with Crank–Nicholson scheme
  • Operator-splitting could be allpied: an elliptic PDE, a parabolic PDE and a set of nonlinear ODEs
  • With small error tolerance, the difference between coupled and decoupled approaches is negligible
  • the main computational burden can be attributed to the solution of the elliptic problem and the set of ODEs (i.e. the ionic model)
  • ODEs are embarrassingly parallel, while parabolic PDEs could also be solved in parallel.
  • The elliptic PDE is the most challenging problem: algebraic multigrid preconditioner (AMG) in conjunction with an iterative Krylov solver

Adaptivity and domain decomposition techniques

  • high spatial and temporal resolution is needed only in the immediate vicinity of a propagating wavefront
  • simpler domain composition methods can be employed, where the region of interest is divided into several subdomains; each subdomain is then integrated at a different rate. Using different temporal resolutions on a per-domain base inevitably leads to load balancing issues for codes executed in parallel

Review of ionic model computation techniques

Integration in standard form

  • Explicit methods (RK4) vs implicit backward methods (e.g. Rosenbrock methods)

Component-wise integration

  • It is common to split the vector formulation into a set of equations, where each ODE is integrated separately.
  • Many of the ODEs comprising an ionic model, typically all gating equations in Hodgkin–Huxley-type models, but also ODEs in Markov state formulations. The Rush-Larsen method (exponential integrators) is popular.
  • Computational cost of log and exp function could be replaced by look-up tables (space vs time)

Tissue-level aspects of ionic model computation

  • Adaptivity in time-stepping may not that useful (load balancing)
  • Careful layout of data structures
  • The maximum time-step that allows stable integration could be limited by the mesh ratio and not the ODEs
  • The potential of expensive ODE integration schemes, which allow large time-steps, cannot be fully exploited since PDE time-step constraints limit Δt to values for which cheaper ODE integrators perform well without any stability problems.
  • increasingly inaccurate solutions with increasing Δt: oscillations in Vm at the end of an action potential upstroke would give rise to artificial calcium transients https://royalsocietypublishing.org/cms/attachment/9bd4936f-c93e-42fb-9bab-b2feccd68d3c/3381fig1.jpg
  • In most cases of practical interest, BDF methods are not likely to help in reducing the computational burden.

A new modelling paradigm for complex ionic models based on temporal multiscale decoupling

The need for a new paradigm

  • Fully implicit methods indeed allow much larger time-steps, but the overall performance was not better and the accuracy was strikingly inferior to that of RL method
  • More realistic model is computationally intensive: Markov process, local control CICR
  • wider range of time-scales: Descriptions of CICR processes stiffen the ODE system considerably ( Δt = 0.2 μs) https://royalsocietypublishing.org/cms/attachment/f4af665b-ec64-4c88-8791-4b7a5e206179/3381fig2.jpg
  • In order to be considered predictive, ionic models must be developed that accurately capture the underlying biophysical mechanisms of experimentally observed phenomena (mechanistic models are preferred)
  • Seperation of time scales: CICR vs Regular vs mitochondria / force
    • not updating all state variables at the same rate (figure 2) leads to significant savings in terms of execution time.

Dynamic reformulation of the ODEs

  • Markov state models, the stiffness of the governing equations may vary substantially depending on the state vector y (the Jacobian is not a constant)
  • the ECME model, stiffness seems to be a problem only for a few microseconds during the onset and upstroke of the calcium transient CICR https://royalsocietypublishing.org/cms/attachment/1399ee9b-591e-4e0a-8b86-8c160828c9f0/3381fig3.jpg
  • Quasi-steady state assumption for subspace calcium

Simulation results

https://user-images.githubusercontent.com/40054455/86706372-6651bd80-c049-11ea-907b-3532ac48b81e.png https://royalsocietypublishing.org/cms/attachment/cafd3caa-5e92-473e-8f24-345a33a48cfe/3381fig4.jpg

Reference


  1. Plank G, Zhou L, Greenstein JL, et al. From mitochondrial ion channels to arrhythmias in the heart: computational techniques to bridge the spatio-temporal scales. Philos. Transact. A Math. Phys. Eng. Sci. 2008;366(1879):3381-3409. doi:10.1098/rsta.2008.0112. PMC2778066↩︎