Computational models of normal and abnormal action potential propagation in cardiac tissue: linking experimental and clinical cardiology

A novel modular modeling approach for understanding different electromechanics between left and right heart in rat

While ion channels and transporters involved in excitation-contraction coupling have been linked and constructed as comprehensive computational models, validation of whether each individual component of a model can be reused has not been previously attempted. Here we address this issue while using a novel modular modeling approach to investigate the underlying mechanism for the differences between left ventricle (LV) and right ventricle (RV). Our model was developed from modules constructed using the module assembly principles of the CellML model markup language. The components of three existing separate models of cardiac function were disassembled as to create smaller modules, validated individually, and then the component parts were combined into a new integrative model of a rat ventricular myocyte. The model was implemented in OpenCOR using the CellML standard in order to ensure reproducibility. Simulated action potential (AP), Ca²⁺ transient, and tension were in close agreement with our experimental measurements: LV AP showed a prolonged duration and a more prominent plateau compared with RV AP; Ca²⁺ transient showed prolonged duration and slow decay in LV compared to RV; the peak value and relaxation of tension were larger and slower, respectively, in LV compared to RV. Our novel approach of module-based mathematical modeling has established that the ionic mechanisms underlying the APs and Ca²⁺ handling play a role in the variation in force production between ventricles. This simulation process also provides a useful way to reuse and elaborate upon existing models in order to develop a new model.

Simple model for identifying critical regions in atrial fibrillation

Atrial fibrillation (AF) is the most common abnormal heart rhythm and the single biggest cause of stroke. Ablation, destroying regions of the atria, is applied largely empirically and can be curative but with a disappointing clinical success rate. We design a simple model of activation wave front propagation on an anisotropic structure mimicking the branching network of heart muscle cells. This integration of phenomenological dynamics and pertinent structure shows how AF emerges spontaneously when the transverse cell-to-cell coupling decreases, as occurs with age, beyond a threshold value. We identify critical regions responsible for the initiation and maintenance of AF, the ablation of which terminates AF. The simplicity of the model allows us to calculate analytically the risk of arrhythmia and express the threshold value of transversal cell-to-cell coupling as a function of the model parameters. This threshold value decreases with increasing refractory period by reducing the number of critical regions which can initiate and sustain microreentrant circuits. These biologically testable predictions might inform ablation therapies and arrhythmic risk assessment.

Construction of 3D MR image-based computer models of pathologic hearts, augmented with histology and optical fluorescence imaging to characterize action potential propagation

Cardiac computer models can help us understand and predict the propagation of excitation waves (i.e., action potential, AP) in healthy and pathologic hearts. Our broad aim is to develop accurate 3D MR image-based computer models of electrophysiology in large hearts (translatable to clinical applications) and to validate them experimentally. The specific goals of this paper were to match models with maps of the propagation of optical AP on the epicardial surface using large porcine hearts with scars, estimating several parameters relevant to macroscopic reaction-diffusion electrophysiological models. We used voltage-sensitive dyes to image AP in large porcine hearts with scars (three specimens had chronic myocardial infarct, and three had radiofrequency RF acute scars). We first analyzed the main AP waves' characteristics: duration (APD) and propagation under controlled pacing locations and frequencies as recorded from 2D optical images. We further built 3D MR image-based computer models that have information derived from the optical measures, as well as morphologic MRI data (i.e., myocardial anatomy, fiber directions and scar definition). The scar morphology from MR images was validated against corresponding whole-mount histology. We also compared the measured 3D isochronal maps of depolarization to simulated isochrones (the latter replicating precisely the experimental conditions), performing model customization and 3D volumetric adjustments of the local conductivity. Our results demonstrated that mean APD in the border zone (BZ) of the infarct scars was reduced by ~13% (compared to ~318 ms measured in normal zone, NZ), but APD did not change significantly in the thin BZ of the ablation scars. A generic value for velocity ratio (1:2.7) in healthy myocardial tissue was derived from measured values of transverse and longitudinal conduction velocities relative to fibers direction (22 cm/s and 60 cm/s, respectively). The model customization and 3D volumetric adjustment reduced the differences between measurements and simulations; for example, from one pacing location, the adjustment reduced the absolute error in local depolarization times by a factor of 5 (i.e., from 58 ms to 11 ms) in the infarcted heart, and by a factor of 6 (i.e., from 60 ms to 9 ms) in the heart with the RF scar. Moreover, the sensitivity of adjusted conductivity maps to different pacing locations was tested, and the errors in activation times were found to be of approximately 10-12 ms independent of pacing location used to adjust model parameters, suggesting that any location can be used for model predictions.

Verification of cardiac tissue electrophysiology simulators using an N-version benchmark

Ongoing developments in cardiac modelling have resulted, in particular, in the development of advanced and increasingly complex computational frameworks for simulating cardiac tissue electrophysiology. The goal of these simulations is often to represent the detailed physiology and pathologies of the heart using codes that exploit the computational potential of high-performance computing architectures. These developments have rapidly progressed the simulation capacity of cardiac virtual physiological human style models; however, they have also made it increasingly challenging to verify that a given code provides a faithful representation of the purported governing equations and corresponding solution techniques. This study provides the first cardiac tissue electrophysiology simulation benchmark to allow these codes to be verified. The benchmark was successfully evaluated on 11 simulation platforms to generate a consensus gold-standard converged solution. The benchmark definition in combination with the gold-standard solution can now be used to verify new simulation codes and numerical methods in the future.

Implications of the Turing completeness of reaction-diffusion models, informed by GPGPU simulations on an XBox 360: cardiac arrhythmias, re-entry and the Halting problem

In the arsenal of tools that a computational modeller can bring to bare on the study of cardiac arrhythmias, the most widely used and arguably the most successful is that of an excitable medium, a special case of a reaction-diffusion model. These are used to simulate the internal chemical reactions of a cardiac cell and the diffusion of their membrane voltages. Via a number of different methodologies it has previously been shown that reaction-diffusion systems are at multiple levels Turing complete. That is, they are capable of computation in the same manner as a universal Turing machine. However, all such computational systems are subject to a limitation known as the Halting problem. By constructing a universal logic gate using a cardiac cell model, we highlight how the Halting problem therefore could limit what it is possible to predict about cardiac tissue, arrhythmias and re-entry. All simulations for this work were carried out on the GPU of an XBox 360 development console, and we also highlight the great gains in computational power and efficiency produced by such general purpose processing on a GPU for cardiac simulations.

Early afterdepolarisations and ventricular arrhythmias in cardiac tissue: a computational study

Afterdepolarisations are associated with arrhythmias in the heart, but are difficult to study experimentally. In this study we used a simplified computational model of 1D and 2D cardiac ventricular tissue, where we could control the size of the region generating afterdepolarisations, as well as the properties of the afterdepolarisation waveform. Provided the size of the afterdepolarisation region was greater than around 1 mm, propagating extrasystoles were produced in both 1D and 2D. The number of extrasystoles produced depended on the amplitude, period, and duration of the oscillatory EAD waveform. In 2D, re-entry was also initiated for specific combinations of EAD amplitude, period, and duration, with the afterdepolarisation region acting as a common pathway. The main finding from this modelling study is therefore that afterdepolarisations can act as potent sources of propagating extrasystoles, as well as a source of re-entrant activation.

A two dimension model of the uterine electrical wave propagation

The uterus, usually quiescent during pregnancy, exhibits forceful contractions at term leading to delivery. These contractions are caused by the synchronized propagation of electrical waves from the pacemaker cells to its neighbors inducing the whole coordinated contraction of the uterus wall leading to labor. In a previous work, we simulate the electrical activity of a single uterine cell by a set of ordinary differential equations. Then, this model has been used to simulate the electrical activity propagation. In the present work, the uterine cell tissue is assumed to have uniform and isotropic propagation, and constant electrical membrane properties. The stability of the numerical solution imposes the choice of a critical temporal step. A wave starts at a pacemaker cell; this electrical activity is initiated by the injection of an external stimulation current to the cell membrane. We observe synchronous wave propagation for axial resistance values around 0.5 GOmega or less and propoagation blocking for values greater than 0.7 GOmega. We compute the conduction velocity of the excitation, for different axial resistance values, and obtain a velocity about 10 cm/sec, approaching the one described by the literature for the rat at end of term.

Asymptotic analysis and analytical solutions of a model of cardiac excitation

We describe an asymptotic approach to gated ionic models of single-cell cardiac excitability. It has a form essentially different from the Tikhonov fast-slow form assumed in standard asymptotic reductions of excitable systems. This is of interest since the standard approaches have been previously found inadequate to describe phenomena such as the dissipation of cardiac wave fronts and the shape of action potential at repolarization. The proposed asymptotic description overcomes these deficiencies by allowing, among other non-Tikhonov features, that a dynamical variable may change its character from fast to slow within a single solution. The general asymptotic approach is best demonstrated on an example which should be both simple and generic. The classical model of Purkinje fibers (Noble in J. Physiol. 160:317-352, 1962) has the simplest functional form of all cardiac models but according to the current understanding it assigns a physiologically incorrect role to the Na current. This leads us to suggest an "Archetypal Model" with the simplicity of the Noble model but with a structure more typical to contemporary cardiac models. We demonstrate that the Archetypal Model admits a complete asymptotic solution in quadratures. To validate our asymptotic approach, we proceed to consider an exactly solvable "caricature" of the Archetypal Model and demonstrate that the asymptotic of its exact solution coincides with the solutions obtained by substituting the "caricature" right-hand sides into the asymptotic solution of the generic Archetypal Model. This is necessary, because, unlike in standard asymptotic descriptions, no general results exist which can guarantee the proximity of the non-Tikhonov asymptotic solutions to the solutions of the corresponding detailed ionic model.

Regional localisation of left ventricular sheet structure: integration with current models of cardiac fibre, sheet and band structure

The architecture of the heart remains controversial despite extensive effort and recent advances in imaging techniques. Several opposing and non-mutually compatible models have been proposed to explain cardiac structure, and these models, although limited, have advanced the study and understanding of heart structure, function and development. We describe key areas of similarity and difference, highlight areas of contention and point to the important limitations of these models. Recent research in animal models on the nature, geometry and interaction of cardiac sheet structure allows unification of some seemingly conflicting features of the structural models. Intriguingly, evidence points to significant inter-individual structural variability (within constrained limits) in the canine, leading to the concept of a continuum (or distribution) of cardiac structures. This variability in heart structure partly explains the ongoing debate on myocardial architecture. These developments are used to construct an integrated description of cardiac structure unifying features of fibre, sheet and band architecture that provides a basis for (i) explaining cardiac electromechanics, (ii) computational simulations of cardiac physiology and (iii) designing interventions.

Conditions for propagation and block of excitation in an asymptotic model of atrial tissue

Detailed ionic models of cardiac cells are difficult for numerical simulations because they consist of a large number of equations and contain small parameters. The presence of small parameters, however, may be used for asymptotic reduction of the models. Earlier results have shown that the asymptotics of cardiac equations are nonstandard. Here we apply such a novel asymptotic method to an ionic model of human atrial tissue to obtain a reduced but accurate model for the description of excitation fronts. Numerical simulations of spiral waves in atrial tissue show that wave fronts of propagating action potentials break up and self-terminate. Our model, in particular, yields a simple analytical criterion of propagation block, which is similar in purpose but completely different in nature to the "Maxwell rule" in the FitzHugh-Nagumo type models. Our new criterion agrees with direct numerical simulations of breakup of reentrant waves.

Dispersion of cardiac action potential duration and the initiation of re-entry: a computational study

BACKGROUND

The initiation of re-entrant cardiac arrhythmias is associated with increased dispersion of repolarisation, but the details are difficult to investigate either experimentally or clinically. We used a computational model of cardiac tissue to study systematically the association between action potential duration (APD) dispersion and susceptibility to re-entry.

METHODS

We simulated a 60 x 60 mm2 D sheet of cardiac ventricular tissue using the Luo-Rudy phase 1 model, with maximal conductance of the K+ channel gKmax set to 0.004 mS mm(-2). Within the central 40 x 40 mm region we introduced square regions with prolonged APD by reducing gKmax to between 0.001 and 0.003 mS mm(-2). We varied (i) the spatial scale of these regions, (ii) the magnitude of gKmax in these regions, and (iii) cell-to-cell coupling.

RESULTS

Changing spatial scale from 5 to 20 mm increased APD dispersion from 49 to 102 ms, and the susceptible window from 31 to 86 ms. Decreasing gKmax in regions with prolonged APD from 0.003 to 0.001 mS mm-2 increased APD dispersion from 22 to 70 ms, and the susceptible window from <1 to 56 ms. Decreasing cell-to-cell coupling by changing the diffusion coefficient from 0.2 to 0.05 mm2 ms(-1) increased APD dispersion from 57 to 88 ms, and increased the susceptible window from 41 to 74 ms.

CONCLUSION

We found a close association between increased APD dispersion and susceptibility to re-entrant arrhythmias, when APD dispersion is increased by larger spatial scale of heterogeneity, greater electrophysiological heterogeneity, and weaker cell-to-cell coupling.

Propagation of normal beats and re-entry in a computational model of ventricular cardiac tissue with regional differences in action potential shape and duration

There is substantial experimental evidence from studies using both intact tissue and isolated single cells to support the existence of different cell types within the ventricular wall of the heart, each possessing different electrical properties. However other studies have failed to find these differences, and instead support the idea that electrical coupling in vivo between regions with different cell types smoothes out differences in action potential shape and duration. In this study we have used a computational model of electrical activation in heterogenous 2D and 3D cardiac tissue to investigate the propagation of both normal beats and arrhythmias. We used the Luo-Rudy dynamic model for guinea pig ventricular cells, with simplified Ca2+ handling and transmural heterogeneity in IKs and Ito. With normal cell-to-cell coupling, a layer of M cells was not necessary for the formation of an upright T wave in the simulated electrocardiogram, and the amplitude and configuration of the T wave was not greatly affected by the thickness and configuration of the M cell layer. Transmural gradients in repolarisation pushed re-entrant waves with an intramural filament towards either the base or the apex of the ventricles, and caused transient break up of re-entry with a transmural filament.

Filament behavior in a computational model of ventricular fibrillation in the canine heart

The aim of this paper was to quantify the behavior of filaments in a computational model of re-entrant ventricular fibrillation. We simulated cardiac activation in an anisotropic monodomain with excitation described by the Fenton-Karma model with Beeler-Reuter restitution, and geometry by the Auckland canine ventricle. We initiated re-entry in the left and right ventricular free walls, as well as the septum. The number of filaments increased during the first 1.5 s before reaching a plateau with a mean value of about 36 in each simulation. Most re-entrant filaments were between 10 and 20 mm long. The proportion of filaments touching the epicardial surface was 65%, but most of these were visible for much less than one period of re-entry. This paper shows that useful information about filament dynamics can be gleaned from models of fibrillation in complex geometries, and suggests that the interplay of filament creation and destruction may offer a target for antifibrillatory therapy.

Effect of regional differences in cardiac cellular electrophysiology on the stability of ventricular arrhythmias: a computational study

Re-entry is an important mechanism of cardiac arrhythmias. During re-entry a wave of electrical activation repeatedly propagates into recovered tissue, rotating around a rod-like filament. Breakdown of a single re-entrant wave into multiple waves is believed to underlie the transition from ventricular tachycardia to ventricular fibrillation. Several mechanisms of breakup have been identified including the effect of anisotropic conduction in the ventricular wall. Cells in the inner and outer layers of the ventricular wall have different action potential durations (APD), and support re-entrant waves with different periods. The aim of this study was to use a computational approach to study twisting and breakdown in a transmural re-entrant wave spanning these regions, and examine the relative role of this effect and anisotropic conduction. We used a simplified model of action potential conduction in the ventricular wall that we modified so that it supported stable re-entry in an anisotropic model with uniform APD. We first examined the effect of regional differences on breakdown in an isotropic model with transmural differences in APD, and found that twisting of the re-entrant filament resulted in buckling and breakdown during the second cycle of re-entry. We found that breakdown was amplified in the anisotropic model, resulting in complex activation in the region of longest APD. This study shows that regional differences in cardiac electrophysiology are a potentially important mechanism for destabilizing re-entry and may act synergistically with other mechanisms to mediate the transition from ventricular tachycardia to ventricular fibrillation.

Mathematical modelling for the new millenium: medicine by numbers

Physicists, engineers and mathematicians are accustomed to the combination of elegance, rigour and utility that characterise mathematical models. They are familiar with the need to dip into their mathematical toolbox to select the technique of choice. However, medicine and biology have not been characterised, in general, by a mathematical formalism. The relative paucity of mathematical models in biology and medicine reflects in part the difficulty in making accurate and appropriate experimental measurements in the field. Signal noise, the lack of appropriate sensors, and uncertainty as to what constitutes the significant measurements are largely to blame for this. The purpose of this paper is to characterise a 'good' model, encourage the development and application of such models to new areas, and outline future developments in the field. It is proposed that a good model will be accurate, predictive, economical, unique and elegant. These principles will be illustrated with reference to four models: radiosensitisation of tumours, modelling solute clearance in haemodialysis, the myogenic response in reactive hyperaemia and cardiac electrical activity. It is suggested that, in the immediate future, the mathematical model will become a useful adjunct to laboratory experiment (and possibly clinical trial), and the provision of 'in silico' models will become routine.

Computational framework for simulating the mechanisms and ECG of re-entrant ventricular fibrillation

Ventricular arrhythmias remain an important cause of morbidity and mortality in the Western world. Although the underlying mechanisms of these arrhythmias can be studied experimentally, these investigations are in general limited to mapping electrical activity on the heart surface. Computational models of action potential propagation offer a potentially powerful way to study electrical activation and arrhythmias, but current models are not easy to link to the clinical environment. In this paper, we describe a framework for computing action potential propagation in which the geometry, electrophysiology and regional properties of ventricular myocardium can be specified so that, for example, different models for cardiac cellular electrophysiology can be used. We have computed action potential propagation during both normal beats and re-entry in an anatomically accurate model of ventricular geometry. By computing the resultant electric current flow in the torso we have also generated simulated ECG signals that result from specific activation patterns in the ventricular model. Models can be powerful tools for explaining observations, and this approach is able to provide a direct link between the different configurations of re-entry observed in computational and experimental studies, and the ECG signals observed in patients.

Modified ionic models of cardiac tissue for efficient large scale computations

Recirculation of excitation, or re-entry, is one of the most important mechanisms of life-threatening cardiac arrhythmias and fibrillation. Modelling these phenomena requires large scale computations in two and three-dimensional slabs of cardiac tissue. Because of computational constraints, most of the studies use simplified (non-ionic) models of cardiac tissue, which are electrophysiologically less accurate than the detailed ionic models. In this paper, we propose a method to modify ionic models of cardiac tissue into an intermediate class of models, which are almost as efficient for computations as simplified models, and retain most of the properties of the original ionic models, such as the shape of the action potential, the restitution of action potential duration and of the conduction velocity, as well as unchanged description of most of the ionic currents.

Dynamics and interaction of filaments in a computational model of re-entrant ventricular fibrillation

Ventricular fibrillation (VF) is a lethal cardiac arrhythmia. Re-entry, in which action potential wavefronts rotate around filaments, is believed to sustain VF. In this study we used a computational model of multiple wavelet fibrillation in the thin-walled right ventricle (10 mm thick) and the thicker walled left ventricle (16 mm thick) to investigate the effect of tissue thickness and initiation protocol on re-entry, and to examine whether filament dynamics and interaction in the model could explain why re-entry is both rarely observed and short-lived in experimental studies that map electrical activation on the heart surface. We found (i) that the density of filaments, the proportion of transmural filaments and the proportion of filaments visible on the model surface were all higher in the 10 mm simulation, (ii) that the initiation protocol influences the rate of filament breakdown but not the number of filaments present after 1 s, and (iii) that although many filaments are visible on the surface of the model, the majority are visible for less than one rotation. This study shows that tissue thickness, geometry and initiation protocol influence electrical activation during VF, and that the rapid motion and interaction of filaments result in transient appearance of surface re-entry.