The cell model¶
The shipped cell models, in examples/odefiles/, join two published models in one
gotranx .ode file:
ToR-ORd (Tomek et al., 2019), a model of the human ventricular action potential and calcium handling, in its dynamic-chloride variant for endocardial cells;
Land (Land et al., 2017), a model of human cardiac contraction.
In the Land model, cytosolic calcium binds to troponin (CaTrpn), and the fraction of
blocked tropomyosin (TmB) falls. Crossbridges then form, first weakly bound (XW) and
then strongly bound (XS). Two distortion states, Zetas and Zetaw, track how far the
attached crossbridges are strained by the stretch rate of the fibre:
The active tension is
where is the length dependence and (120 kPa by default) the reference tension. The tension therefore depends on the stretch rate, and this velocity dependence is what makes the coupling scheme matter (Coupling).
Splits¶
Every .ode file marks one component "mechanics". Its states are stepped by the
activation backend at the quadrature points of the mechanics mesh, and everything else
is stepped by EP. simcardemsx.ode_model.load_ode_modules(odefile, output_dir) cuts the
file there and generates two modules, both with gotranx’s generalized Rush-Larsen
scheme:
ep_model.py, NumPy code, forbeat.odesolver.DolfinODESolver;mechanics_model.py, the same scheme printed as UFL expressions, so the step can be evaluated symbolically on the mechanics mesh and differentiated by Newton.
The generated code is cached in output_dir, with a stamp of the .ode file’s hash,
gotranx’s version and the generator’s own source. It is regenerated only when one of
them changes.
The three shipped splits of ToR-ORd + Land:
| File | Mechanics states | EP → mechanics | Mechanics → EP |
|---|---|---|---|
…_zetasplit.ode | Zetas, Zetaw, Cd | XS, XW | Zetas, Zetaw, |
…_catrpnsplit.ode | the Land model except troponin | CaTrpn | |
…_caisplit.ode | the whole Land model | cai | J_TRPN |
The further the cut moves towards EP, the more of the contraction model is evaluated at
the true stretch of the mechanics mesh. On the Ca split, EP sees the contraction
only through the troponin buffering flux J_TRPN. That split is also the one the NumPy
crossbridge models of CrossbridgeSegregated plug into.
Tutorial 1 runs the whole cell, and then each split.
Writing your own split¶
Any gotranx .ode file works if:
one component is named
"mechanics";that component has the parameters
lmbda(fibre stretch) anddLambda(stretch rate), which the backend sets from the displacement;it has a monitor for the tension (
Taby default, in kPa, or settension=andtension_unit=onGeneratedActivation).
Active stress¶
The tension acts along the fibre direction . By default (formulation="stretch")
simcardemsx uses the normalization of Regazzoni & Quarteroni (2021),
under which the tension is the force per reference area along the fibre whatever the
stretch. The historical simcardems form, formulation="invariant"
(), is larger by a factor of . It is
kept only to reproduce old results, which are not comparable with new ones.
tension_scale= multiplies in the stress and nothing else. A DG0 mask of 0 and 1
takes the active stress out of part of the domain, for example the valve plugs of the
rodero_05 example, while the contraction states there keep their own dynamics.
Passive mechanics¶
The passive material, compressibility and viscosity are pulse’s. The examples use the Holzapfel-Ogden model (Holzapfel & Ogden, 2009), either incompressible or with pulse’s compressibility penalties. See the pulse documentation for the full list.
References¶
J. Tomek et al. Development, calibration, and validation of a novel human ventricular myocyte model in health, disease, and drug block. eLife 8, e48890 (2019).
S. Land et al. A model of cardiac contraction based on novel measurements of tension development in human cardiomyocytes. J. Mol. Cell. Cardiol. 106, 68–83 (2017).
G. A. Holzapfel and R. W. Ogden. Constitutive modelling of passive myocardium: a structurally based framework for material characterization. Phil. Trans. R. Soc. A 367, 3445–3475 (2009).
F. Regazzoni and A. Quarteroni. An oscillation-free fully staggered algorithm for velocity-dependent active models of cardiac mechanics. Comput. Methods Appl. Mech. Eng. 373, 113506 (2021).