Dual effects of the small-conductance Ca2+-activated K+ current on human atrial electrophysiology and Ca2+-driven arrhythmogenesis: an in silico study.

By sensing changes in intracellular Ca2+, small-conductance Ca2+-activated K+ (SK) channels dynamically regulate the dynamics of the cardiac action potential (AP) on a beat-to-beat basis. Given their predominance in atria versus ventricles, SK channels are considered a promising atrial-selective pharmacological target against atrial fibrillation (AF), the most common cardiac arrhythmia. However, the precise contribution of SK current (ISK) to atrial arrhythmogenesis is poorly understood, and may potentially involve different mechanisms that depend on species, heart rates, and degree of AF-induced atrial remodeling. Both reduced and enhanced ISK have been linked to AF. Similarly, both SK channel up- and downregulation have been reported in chronic AF (cAF) versus normal sinus rhythm (nSR) patient samples. Here, we use our multiscale modeling framework to obtain mechanistic insights into the contribution of ISK in human atrial cardiomyocyte electrophysiology. We simulate several protocols to quantify how ISK modulation affects the regulation of AP duration (APD), Ca2+ transient, refractoriness, and occurrence of alternans and delayed afterdepolarizations (DADs). Our simulations show that ISK activation shortens the APD and atrial effective refractory period, limits Ca2+ cycling, and slightly increases the propensity for alternans in both nSR and cAF conditions. We also show that increasing ISK counteracts DAD development by enhancing the repolarization force that opposes the Ca2+-dependent depolarization. Taken together, our results suggest that increasing ISK in human atrial cardiomyocytes could promote reentry while protecting against triggered activity. Depending on the leading arrhythmogenic mechanism, ISK inhibition may thus be a beneficial or detrimental anti-AF strategy.NEW & NOTEWORTHY Using our established framework for human atrial myocyte simulations, we investigated the role of the small-conductance Ca2+-activated K+ current (ISK) in the regulation of cell function and the development of Ca2+-driven arrhythmias. We found that ISK inhibition, a promising atrial-selective pharmacological strategy against atrial fibrillation, counteracts the reentry-promoting abbreviation of atrial refractoriness, but renders human atrial myocytes more vulnerable to delayed afterdepolarizations, thus potentially increasing the propensity for ectopic (triggered) activity.

A deep generative model of 3D single-cell organization.

We introduce a framework for end-to-end integrative modeling of 3D single-cell multi-channel fluorescent image data of diverse subcellular structures. We employ stacked conditional ß-variational autoencoders to first learn a latent representation of cell morphology, and then learn a latent representation of subcellular structure localization which is conditioned on the learned cell morphology. Our model is flexible and can be trained on images of arbitrary subcellular structures and at varying degrees of sparsity and reconstruction fidelity. We train our full model on 3D cell image data and explore design trade-offs in the 2D setting. Once trained, our model can be used to predict plausible locations of structures in cells where these structures were not imaged. The trained model can also be used to quantify the variation in the location of subcellular structures by generating plausible instantiations of each structure in arbitrary cell geometries. We apply our trained model to a small drug perturbation screen to demonstrate its applicability to new data. We show how the latent representations of drugged cells differ from unperturbed cells as expected by on-target effects of the drugs.

Label-free prediction of three-dimensional fluorescence images from transmitted-light microscopy.

Understanding cells as integrated systems is central to modern biology. Although fluorescence microscopy can resolve subcellular structure in living cells, it is expensive, is slow, and can damage cells. We present a label-free method for predicting three-dimensional fluorescence directly from transmitted-light images and demonstrate that it can be used to generate multi-structure, integrated images. The method can also predict immunofluorescence (IF) from electron micrograph (EM) inputs, extending the potential applications.

DLITE Uses Cell-Cell Interface Movement to Better Infer Cell-Cell Tensions.

Cell shapes and connectivities evolve over time as the colony changes shape or embryos develop. Shapes of intercellular interfaces are closely coupled with the forces resulting from actomyosin interactions, membrane tension, or cell-cell adhesions. Although it is possible to computationally infer cell-cell forces from a mechanical model of collective cell behavior, doing so for temporally evolving forces in a manner robust to digitization difficulties is challenging. Here, we introduce a method for dynamic local intercellular tension estimation (DLITE) that infers such evolution in temporal force with less sensitivity to digitization ambiguities or errors. This method builds upon previous work on single time points (cellular force-inference toolkit). We validate our method using synthetic geometries. DLITE's inferred cell colony tension evolutions correlate better with ground truth for these synthetic geometries as compared to tension values inferred from methods that consider each time point in isolation. We introduce cell connectivity errors, angle estimate errors, connection mislocalization, and connection topological changes to synthetic data and show that DLITE has reduced sensitivity to these conditions. Finally, we apply DLITE to time series of human-induced pluripotent stem cell colonies with endogenously expressed GFP-tagged zonulae occludentes-1. We show that DLITE offers improved stability in the inference of cell-cell tensions and supports a correlation between the dynamics of cell-cell forces and colony rearrangement.

A computational framework for testing arrhythmia marker sensitivities to model parameters in functionally calibrated populations of atrial cells.

Models of cardiac cell electrophysiology are complex non-linear systems which can be used to gain insight into mechanisms of cardiac dynamics in both healthy and pathological conditions. However, the complexity of cardiac models can make mechanistic insight difficult. Moreover, these are typically fitted to averaged experimental data which do not incorporate the variability in observations. Recently, building populations of models to incorporate inter- and intra-subject variability in simulations has been combined with sensitivity analysis (SA) to uncover novel ionic mechanisms and potentially clarify arrhythmogenic behaviors. We used the Koivumäki human atrial cell model to create two populations, representing normal Sinus Rhythm (nSR) and chronic Atrial Fibrillation (cAF), by varying 22 key model parameters. In each population, 14 biomarkers related to the action potential and dynamic restitution were extracted. Populations were calibrated based on distributions of biomarkers to obtain reasonable physiological behavior, and subjected to SA to quantify correlations between model parameters and pro-arrhythmia markers. The two populations showed distinct behaviors under steady state and dynamic pacing. The nSR population revealed greater variability, and more unstable dynamic restitution, as compared to the cAF population, suggesting that simulated cAF remodeling rendered cells more stable to parameter variation and rate adaptation. SA revealed that the biomarkers depended mainly on five ionic currents, with noted differences in sensitivities to these between nSR and cAF. Also, parameters could be selected to produce a model variant with no alternans and unaltered action potential morphology, highlighting that unstable dynamical behavior may be driven by specific cell parameter settings. These results ultimately suggest that arrhythmia maintenance in cAF may not be due to instability in cell membrane excitability, but rather due to tissue-level effects which promote initiation and maintenance of reentrant arrhythmia.

Refractoriness in human atria: Time and voltage dependence of sodium channel availability.

BACKGROUND: Refractoriness of cardiac cells limits maximum frequency of electrical activity and protects the heart from tonic contractions. Short refractory periods support major arrhythmogenic substrates and augmentation of refractoriness is therefore seen as a main mechanism of antiarrhythmic drugs. Cardiomyocyte excitability depends on availability of sodium channels, which involves both time- and voltage-dependent recovery from inactivation. This study therefore aims to characterise how sodium channel inactivation affects refractoriness in human atria. METHODS AND RESULTS: Steady-state activation and inactivation parameters of sodium channels measured in vitro in isolated human atrial cardiomyocytes were used to parameterise a mathematical human atrial cell model. Action potential data were acquired from human atrial trabeculae of patients in either sinus rhythm or chronic atrial fibrillation. The ex vivo measurements of action potential duration, effective refractory period and resting membrane potential were well-replicated in simulations using this new in silico model. Notably, the voltage threshold potential at which refractoriness was observed was not different between sinus rhythm and chronic atrial fibrillation tissues and was neither affected by changes in frequency (1 vs. 3Hz). CONCLUSIONS: Our results suggest a preferentially voltage-dependent, rather than time-dependent, effect with respect to refractoriness at physiologically relevant rates in human atria. However, as the resting membrane potential is hyperpolarized in chronic atrial fibrillation, the voltage-dependence of excitability dominates, profoundly increasing the risk for arrhythmia re-initiation and maintenance in fibrillating atria. Our results thereby highlight resting membrane potential as a potential target in pharmacological management of chronic atrial fibrillation.

In silico screening of the key cellular remodeling targets in chronic atrial fibrillation.

Chronic atrial fibrillation (AF) is a complex disease with underlying changes in electrophysiology, calcium signaling and the structure of atrial myocytes. How these individual remodeling targets and their emergent interactions contribute to cell physiology in chronic AF is not well understood. To approach this problem, we performed in silico experiments in a computational model of the human atrial myocyte. The remodeled function of cellular components was based on a broad literature review of in vitro findings in chronic AF, and these were integrated into the model to define a cohort of virtual cells. Simulation results indicate that while the altered function of calcium and potassium ion channels alone causes a pronounced decrease in action potential duration, remodeling of intracellular calcium handling also has a substantial impact on the chronic AF phenotype. We additionally found that the reduction in amplitude of the calcium transient in chronic AF as compared to normal sinus rhythm is primarily due to the remodeling of calcium channel function, calcium handling and cellular geometry. Finally, we found that decreased electrical resistance of the membrane together with remodeled calcium handling synergistically decreased cellular excitability and the subsequent inducibility of repolarization abnormalities in the human atrial myocyte in chronic AF. We conclude that the presented results highlight the complexity of both intrinsic cellular interactions and emergent properties of human atrial myocytes in chronic AF. Therefore, reversing remodeling for a single remodeled component does little to restore the normal sinus rhythm phenotype. These findings may have important implications for developing novel therapeutic approaches for chronic AF.

Variable t-tubule organization and Ca2+ homeostasis across the atria.

Although t-tubules have traditionally been thought to be absent in atrial cardiomyocytes, recent studies have suggested that t-tubules exist in the atria of large mammals. However, it is unclear whether regional differences in t-tubule organization exist that define cardiomyocyte function across the atria. We sought to investigate regional t-tubule density in pig and rat atria and the consequences for cardiomyocyte Ca(2+) homeostasis. We observed t-tubules in approximately one-third of rat atrial cardiomyocytes, in both tissue cryosections and isolated cardiomyocytes. In a minority (≈10%) of atrial cardiomyocytes, the t-tubular network was well organized, with a transverse structure resembling that of ventricular cardiomyocytes. In both rat and pig atrial tissue, we observed higher t-tubule density in the epicardium than in the endocardium. Consistent with high variability in the distribution of t-tubules and Ca(2+) channels among cells, L-type Ca(2+) current amplitude was also highly variable and steeply dependent on capacitance and t-tubule density. Accordingly, Ca(2+) transients showed great variability in Ca(2+) release synchrony. Simultaneous imaging of the cell membrane and Ca(2+) transients confirmed t-tubule functionality. Results from mathematical modeling indicated that a transmural gradient in t-tubule organization and Ca(2+) release kinetics supports synchronization of contraction across the atrial wall and may underlie transmural differences in the refractory period. In conclusion, our results indicate that t-tubule density is highly variable across the atria. We propose that higher t-tubule density in cells localized in the epicardium may promote synchronization of contraction across the atrial wall.

NS5806 partially restores action potential duration but fails to ameliorate calcium transient dysfunction in a computational model of canine heart failure.

AIMS: The study investigates how increased Ito, as mediated by the activator NS5806, affects excitation-contraction coupling in chronic heart failure (HF). We hypothesized that restoring spike-and-dome morphology of the action potential (AP) to a healthy phenotype would be insufficient to restore the intracellular Ca(2) (+) transient (CaT), due to HF-induced remodelling of Ca(2+) handling. METHODS AND RESULTS: An existing mathematical model of the canine ventricular myocyte was modified to incorporate recent experimental data from healthy and failing myocytes, resulting in models of both healthy and HF epicardial, midmyocardial, and endocardial cell variants. Affects of NS5806 were also included in HF models through its direct interaction with Kv4.3 and Kv1.4. Single-cell simulations performed in all models (control, HF, and HF + drug) and variants (epi, mid, and endo) assessed AP morphology and underlying ionic processes with a focus on calcium transients (CaT), how these were altered in HF across the ventricular wall, and the subsequent effects of varying compound concentration in HF. Heart failure model variants recapitulated a characteristic increase in AP duration (APD) in the disease. The qualitative effects of application of half-maximal effective concentration (EC50) of NS5806 on APs and CaT are heterogeneous and non-linear. Deepening in the AP notch with drug is a direct effect of the activation of Ito; both Ito and consequent alteration of IK1 kinetics cause decrease in AP plateau potential. Decreased APD50 and APD90 are both due to altered IK1. Analysis revealed that drug effects depend on transmurality. Ca(2+) transient morphology changes-increased amplitude and shorter time to peak-are due to direct increase in ICa,L and indirect larger SR Ca(2+) release subsequent to Ito activation. CONCLUSIONS: Downstream effects of a compound acting exclusively on sarcolemmal ion channels are difficult to predict. Remediation of APD to pre-failing state does not ameliorate dysfunction in CaT; however, restoration of notch depth appears to impart modest benefit and a likelihood of therapeutic value in modulating early repolarization.

Fibrosis modeling choice affects morphology of ventricular arrhythmia in non-ischemic cardiomyopathy.

Introduction: Patients with non-ischemic cardiomyopathy (NICM) are at risk for ventricular arrhythmias, but diagnosis and treatment planning remain a serious clinical challenge. Although computational modeling has provided valuable insight into arrhythmic mechanisms, the optimal method for simulating reentry in NICM patients with structural disease is unknown. Methods: Here, we compare the effects of fibrotic representation on both reentry initiation and reentry morphology in patient-specific cardiac models. We investigate models with heterogeneous networks of non-conducting structures (cleft models) and models where fibrosis is represented as a dense core with a surrounding border zone (non-cleft models). Using segmented cardiac magnetic resonance with late gadolinium enhancement (LGE) of five NICM patients, we created 185 3D ventricular electrophysiological models with different fibrotic representations (clefts, reduced conductivity and ionic remodeling). Results: Reentry was induced by electrical pacing in 647 out of 3,145 simulations. Both cleft and non-cleft models can give rise to double-loop reentries meandering through fibrotic regions (Type 1-reentry). When accounting for fibrotic volume, the initiation sites of these reentries are associated with high local fibrotic density (mean LGE in cleft models: p< 0.001, core volume in non-cleft models: p = 0.018, negative binomial regression). In non-cleft models, Type 1-reentries required slow conduction in core tissue (non-cleftsc models) as opposed to total conduction block. Incorporating ionic remodeling in fibrotic regions can give rise to single- or double-loop rotors close to healthy-fibrotic interfaces (Type 2-reentry). Increasing the cleft density or core-to-border zone ratio in cleft and non-cleftc models, respectively, leads to increased inducibility and a change in reentry morphology from Type 2 to Type 1. Conclusions: By demonstrating how fibrotic representation affects reentry morphology and location, our findings can aid model selection for simulating arrhythmogenesis in NICM.

Stretch of the papillary insertion triggers reentrant arrhythmia: an in silico patient study.

Background: The electrophysiological mechanism connecting mitral valve prolapse (MVP), premature ventricular complexes and life-threatening ventricular arrhythmia is unknown. A common hypothesis is that stretch activated channels (SACs) play a significant role. SACs can trigger depolarizations or shorten repolarization times in response to myocardial stretch. Through these mechanisms, pathological traction of the papillary muscle (PM), as has been observed in patients with MVP, may induce irregular electrical activity and result in reentrant arrhythmia. Methods: Based on a patient with MVP and mitral annulus disjunction, we modeled the effect of excessive PM traction in a detailed medical image-derived ventricular model by activating SACs in the PM insertion region. By systematically varying the onset of SAC activation following sinus pacing, we identified vulnerability windows for reentry with 1 ms resolution. We explored how reentry was affected by the SAC reversal potential ( E SAC ) and the size of the region with simulated stretch (SAC region). Finally, the effect of global or focal fibrosis, modeled as reduction in tissue conductivity or mesh splitting (fibrotic microstructure), was investigated. Results: In models with healthy tissue or fibrosis modeled solely as CV slowing, we observed two vulnerable periods of reentry: For E SAC of -10 and -30 mV, SAC activated during the T-wave could cause depolarization of the SAC region which lead to reentry. For E SAC of -40 and -70 mV, SAC activated during the QRS complex could result in early repolarization of the SAC region and subsequent reentry. In models with fibrotic microstructure in the SAC region, we observed micro-reentries and a larger variability in which times of SAC activation triggered reentry. In these models, 86% of reentries were triggered during the QRS complex or T-wave. We only observed reentry for sufficiently large SAC regions ( > = 8 mm radius in models with healthy tissue). Conclusion: Stretch of the PM insertion region following sinus activation may initiate ventricular reentry in patients with MVP, with or without fibrosis. Depending on the SAC reversal potential and timing of stretch, reentry may be triggered by ectopy due to SAC-induced depolarizations or by early repolarization within the SAC region.

Note on a possible proarrhythmic property of antiarrhythmic drugs aimed at improving gap-junction coupling.

Reduced conduction velocity (CV) in the myocardium is well known to increase the probability of arrhythmia and can be caused by structural changes, reduced excitability of individual myocytes, or decreased electrical coupling in the tissue. Recently, investigators have developed antiarrhythmic drugs that target the connections between individual myocytes with the goal of restoring tissue CV, specifically through increasing gap-junction coupling. In a simple but qualitatively relevant mathematical model, we show here that the introduction of a drug that improves intercellular conductance will indeed increase the CV. However, conditions that would require such a drug, such as fibrotic remodeling, may also increase the load of fibroblasts. Fibroblasts may couple to myocytes in much the same way as myocytes couple to each other, and therefore the use of such an agent may also improve coupling between myocytes and fibroblasts. We present numerical examples illustrating that when the load of coupled fibroblasts on myocytes is low or nonexistent, the drug works as expected, i.e., the drug increases CV. On the other hand, when the fibroblast load is high, changes in CV are nonmonotonic, i.e., the CV first increases and then decreases with an increase in dosage. The existence of coupled fibroblasts may therefore impair the effect of the drug, and under unfortunate conditions may be proarrhythmic.

Computational cardiac physiology for new modelers: Origins, foundations, and future.

Mathematical models of the cardiovascular system have come a long way since they were first introduced in the early 19th century. Driven by a rapid development of experimental techniques, numerical methods, and computer hardware, detailed models that describe physical scales from the molecular level up to organs and organ systems have been derived and used for physiological research. Mathematical and computational models can be seen as condensed and quantitative formulations of extensive physiological knowledge and are used for formulating and testing hypotheses, interpreting and directing experimental research, and have contributed substantially to our understanding of cardiovascular physiology. However, in spite of the strengths of mathematics to precisely describe complex relationships and the obvious need for the mathematical and computational models to be informed by experimental data, there still exist considerable barriers between experimental and computational physiological research. In this review, we present a historical overview of the development of mathematical and computational models in cardiovascular physiology, including the current state of the art. We further argue why a tighter integration is needed between experimental and computational scientists in physiology, and point out important obstacles and challenges that must be overcome in order to fully realize the synergy of experimental and computational physiological research.

Identifying risk of adverse outcomes in COVID-19 patients via artificial intelligence-powered analysis of 12-lead intake electrocardiogram.

BACKGROUND: Adverse events in COVID-19 are difficult to predict. Risk stratification is encumbered by the need to protect healthcare workers. We hypothesize that artificial intelligence (AI) can help identify subtle signs of myocardial involvement in the 12-lead electrocardiogram (ECG), which could help predict complications. OBJECTIVE: Use intake ECGs from COVID-19 patients to train AI models to predict risk of mortality or major adverse cardiovascular events (MACE). METHODS: We studied intake ECGs from 1448 COVID-19 patients (60.5% male, aged 63.4 ± 16.9 years). Records were labeled by mortality (death vs discharge) or MACE (no events vs arrhythmic, heart failure [HF], or thromboembolic [TE] events), then used to train AI models; these were compared to conventional regression models developed using demographic and comorbidity data. RESULTS: A total of 245 (17.7%) patients died (67.3% male, aged 74.5 ± 14.4 years); 352 (24.4%) experienced at least 1 MACE (119 arrhythmic, 107 HF, 130 TE). AI models predicted mortality and MACE with area under the curve (AUC) values of 0.60 ± 0.05 and 0.55 ± 0.07, respectively; these were comparable to AUC values for conventional models (0.73 ± 0.07 and 0.65 ± 0.10). There were no prominent temporal trends in mortality rate or MACE incidence in our cohort; holdout testing with data from after a cutoff date (June 9, 2020) did not degrade model performance. CONCLUSION: Using intake ECGs alone, our AI models had limited ability to predict hospitalized COVID-19 patients' risk of mortality or MACE. Our models' accuracy was comparable to that of conventional models built using more in-depth information, but translation to clinical use would require higher sensitivity and positive predictive value. In the future, we hope that mixed-input AI models utilizing both ECG and clinical data may be developed to enhance predictive accuracy.

Susceptibility to arrhythmia in the infarcted heart depends on myofibroblast density.

Fibroblasts are electrophysiologically quiescent in the healthy heart. Evidence suggests that remodeling following myocardial infarction may include coupling of myofibroblasts (Mfbs) among themselves and with myocytes via gap junctions. We use a magnetic resonance imaging-based, three-dimensional computational model of the chronically infarcted rabbit ventricles to characterize the arrhythmogenic substrate resulting from Mfb infiltration as a function of Mfb density. Mfbs forming gap junctions were incorporated into both infarct regions, the periinfarct zone (PZ) and the scar; six scenarios were modeled: 0%, 10%, and 30% Mfbs in the PZ, with either 80% or 0% Mfbs in the scar. Ionic current remodeling in PZ was also included. All preparations exhibited elevated resting membrane potential within and near the PZ and action potential duration shortening throughout the ventricles. The unique combination of PZ ionic current remodeling and different degrees of Mfb infiltration in the infarcted ventricles determines susceptibility to arrhythmia. At low densities, Mfbs do not alter arrhythmia propensity; the latter arises predominantly from ionic current remodeling in PZ. At intermediate densities, Mfbs cause additional action potential shortening and exacerbate arrhythmia propensity. At high densities, Mfbs protect against arrhythmia by causing resting depolarization and blocking propagation, thus overcoming the arrhythmogenic effects of PZ ionic current remodeling.

Combined In-silico and Machine Learning Approaches Toward Predicting Arrhythmic Risk in Post-infarction Patients.

Background: Remodeling due to myocardial infarction (MI) significantly increases patient arrhythmic risk. Simulations using patient-specific models have shown promise in predicting personalized risk for arrhythmia. However, these are computationally- and time- intensive, hindering translation to clinical practice. Classical machine learning (ML) algorithms (such as K-nearest neighbors, Gaussian support vector machines, and decision trees) as well as neural network techniques, shown to increase prediction accuracy, can be used to predict occurrence of arrhythmia as predicted by simulations based solely on infarct and ventricular geometry. We present an initial combined image-based patient-specific in silico and machine learning methodology to assess risk for dangerous arrhythmia in post-infarct patients. Furthermore, we aim to demonstrate that simulation-supported data augmentation improves prediction models, combining patient data, computational simulation, and advanced statistical modeling, improving overall accuracy for arrhythmia risk assessment. Methods: MRI-based computational models were constructed from 30 patients 5 days post-MI (the "baseline" population). In order to assess the utility biophysical model-supported data augmentation for improving arrhythmia prediction, we augmented the virtual baseline patient population. Each patient ventricular and ischemic geometry in the baseline population was used to create a subfamily of geometric models, resulting in an expanded set of patient models (the "augmented" population). Arrhythmia induction was attempted via programmed stimulation at 17 sites for each virtual patient corresponding to AHA LV segments and simulation outcome, "arrhythmia," or "no-arrhythmia," were used as ground truth for subsequent statistical prediction (machine learning, ML) models. For each patient geometric model, we measured and used choice data features: the myocardial volume and ischemic volume, as well as the segment-specific myocardial volume and ischemia percentage, as input to ML algorithms. For classical ML techniques (ML), we trained k-nearest neighbors, support vector machine, logistic regression, xgboost, and decision tree models to predict the simulation outcome from these geometric features alone. To explore neural network ML techniques, we trained both a three - and a four-hidden layer multilayer perceptron feed forward neural networks (NN), again predicting simulation outcomes from these geometric features alone. ML and NN models were trained on 70% of randomly selected segments and the remaining 30% was used for validation for both baseline and augmented populations. Results: Stimulation in the baseline population (30 patient models) resulted in reentry in 21.8% of sites tested; in the augmented population (129 total patient models) reentry occurred in 13.0% of sites tested. ML and NN models ranged in mean accuracy from 0.83 to 0.86 for the baseline population, improving to 0.88 to 0.89 in all cases. Conclusion: Machine learning techniques, combined with patient-specific, image-based computational simulations, can provide key clinical insights with high accuracy rapidly and efficiently. In the case of sparse or missing patient data, simulation-supported data augmentation can be employed to further improve predictive results for patient benefit. This work paves the way for using data-driven simulations for prediction of dangerous arrhythmia in MI patients.

DeepFake electrocardiograms using generative adversarial networks are the beginning of the end for privacy issues in medicine.

Recent global developments underscore the prominent role big data have in modern medical science. But privacy issues constitute a prevalent problem for collecting and sharing data between researchers. However, synthetic data generated to represent real data carrying similar information and distribution may alleviate the privacy issue. In this study, we present generative adversarial networks (GANs) capable of generating realistic synthetic DeepFake 10-s 12-lead electrocardiograms (ECGs). We have developed and compared two methods, named WaveGAN* and Pulse2Pulse. We trained the GANs with 7,233 real normal ECGs to produce 121,977 DeepFake normal ECGs. By verifying the ECGs using a commercial ECG interpretation program (MUSE 12SL, GE Healthcare), we demonstrate that the Pulse2Pulse GAN was superior to the WaveGAN* to produce realistic ECGs. ECG intervals and amplitudes were similar between the DeepFake and real ECGs. Although these synthetic ECGs mimic the dataset used for creation, the ECGs are not linked to any individuals and may thus be used freely. The synthetic dataset will be available as open access for researchers at OSF.io and the DeepFake generator available at the Python Package Index (PyPI) for generating synthetic ECGs. In conclusion, we were able to generate realistic synthetic ECGs using generative adversarial neural networks on normal ECGs from two population studies, thereby addressing the relevant privacy issues in medical datasets.

Explaining deep neural networks for knowledge discovery in electrocardiogram analysis.

Deep learning-based tools may annotate and interpret medical data more quickly, consistently, and accurately than medical doctors. However, as medical doctors are ultimately responsible for clinical decision-making, any deep learning-based prediction should be accompanied by an explanation that a human can understand. We present an approach called electrocardiogram gradient class activation map (ECGradCAM), which is used to generate attention maps and explain the reasoning behind deep learning-based decision-making in ECG analysis. Attention maps may be used in the clinic to aid diagnosis, discover new medical knowledge, and identify novel features and characteristics of medical tests. In this paper, we showcase how ECGradCAM attention maps can unmask how a novel deep learning model measures both amplitudes and intervals in 12-lead electrocardiograms, and we show an example of how attention maps may be used to develop novel ECG features.

Computational translation of drug effects from animal experiments to human ventricular myocytes.

Using animal cells and tissues as precise measuring devices for developing new drugs presents a long-standing challenge for the pharmaceutical industry. Despite the very significant resources that continue to be dedicated to animal testing of new compounds, only qualitative results can be obtained. This often results in both false positives and false negatives. Here, we show how the effect of drugs applied to animal ventricular myocytes can be translated, quantitatively, to estimate a number of different effects of the same drug on human cardiomyocytes. We illustrate and validate our methodology by translating, from animal to human, the effect of dofetilide applied to dog cardiomyocytes, the effect of E-4031 applied to zebrafish cardiomyocytes, and, finally, the effect of sotalol applied to rabbit cardiomyocytes. In all cases, the accuracy of our quantitative estimates are demonstrated. Our computations reveal that, in principle, electrophysiological data from testing using animal ventricular myocytes, can give precise, quantitative estimates of the effect of new compounds on human cardiomyocytes.

Physiological Effects of the Electrogenic Current Generated by the Na+/K+ Pump in Mammalian Articular Chondrocytes.

Background: Although the chondrocyte is a nonexcitable cell, there is strong interest in gaining detailed knowledge of its ion pumps, channels, exchangers, and transporters. In combination, these transport mechanisms set the resting potential, regulate cell volume, and strongly modulate responses of the chondrocyte to endocrine agents and physicochemical alterations in the surrounding extracellular microenvironment. Materials and Methods: Mathematical modeling was used to assess the functional roles of energy-requiring active transport, the Na+/K+ pump, in chondrocytes. Results: Our findings illustrate plausible physiological roles for the Na+/K+ pump in regulating the resting membrane potential and suggest ways in which specific molecular components of pump can respond to the unique electrochemical environment of the chondrocyte. Conclusion: This analysis provides a basis for linking chondrocyte electrophysiology to metabolism and yields insights into novel ways of manipulating or regulating responsiveness to external stimuli both under baseline conditions and in chronic diseases such as osteoarthritis.