A rapid electromechanical model to predict reverse remodeling following cardiac resynchronization therapy

Characterization of cardiac resynchronization therapy response through machine learning and personalized models

INTRODUCTION

The characterization and selection of heart failure (HF) patients for cardiac resynchronization therapy (CRT) remain challenging, with around 30% non-responder rate despite following current guidelines. This study aims to propose a novel hybrid approach, integrating machine-learning and personalized models, to identify explainable phenogroups of HF patients and predict their CRT response.

METHODS

The paper proposes the creation of a complete personalized model population based on preoperative CRT patient strain curves. Based on the parameters and features extracted from these personalized models, phenotypes of patients are identified thanks to a clustering algorithm and a random forest classification is provided.

RESULTS

A close match was observed between the 162 experimental and simulated myocardial strain curves, with a mean RMSE of 4.48% (±1.08) for the 162 patients. Five phenogroups of personalized models were identified from the clustering, with response rates ranging from 52% to 94%. The classification results show a mean area under the curves (AUC) of 0.86 ± 0.06 and provided a feature importance analysis with 22 features selected. Results show both regional myocardial contractility (from 22.5% to 33.0%), tissue viability and electrical activation delays importance on CRT response for each HF patient (from 55.8 ms to 88.4 ms).

DISCUSSION

The patient-specific model parameters' analysis provides an explainable interpretation of HF patient phenogroups in relation to physiological mechanisms that seem predictive of the CRT response. These novel combined approaches appear as promising tools to improve understanding of LV mechanical dyssynchrony for HF patient characterization and CRT selection.

Guidelines for mechanistic modeling and analysis in cardiovascular research

Computational, or in silico, models are an effective, noninvasive tool for investigating cardiovascular function. These models can be used in the analysis of experimental and clinical data to identify possible mechanisms of (ab)normal cardiovascular physiology. Recent advances in computing power and data management have led to innovative and complex modeling frameworks that simulate cardiovascular function across multiple scales. While commonly used in multiple disciplines, there is a lack of concise guidelines for the implementation of computer models in cardiovascular research. In line with recent calls for more reproducible research, it is imperative that scientists adhere to credible practices when developing and applying computational models to their research. The goal of this manuscript is to provide a consensus document that identifies best practices for in silico computational modeling in cardiovascular research. These guidelines provide the necessary methods for mechanistic model development, model analysis, and formal model calibration using fundamentals from statistics. We outline rigorous practices for computational, mechanistic modeling in cardiovascular research and discuss its synergistic value to experimental and clinical data.

Multiscale Finite Element Modeling of Left Ventricular Growth in Simulations of Valve Disease

Multiscale models of the cardiovascular system are emerging as effective tools for investigating the mechanisms that drive ventricular growth and remodeling. These models can predict how molecular-level mechanisms impact organ-level structure and function and could provide new insights that help improve patient care. MyoFE is a multiscale computer framework that bridges molecular and organ-level mechanisms in a finite element model of the left ventricle that is coupled with the systemic circulation. In this study, we extend MyoFE to include a growth algorithm, based on volumetric growth theory, to simulate concentric growth (wall thickening/thinning) and eccentric growth (chamber dilation/constriction) in response to valvular diseases. Specifically in our model, concentric growth is controlled by time-averaged total stress along the fiber direction over a cardiac cycle while eccentric growth responds to time-averaged intracellular myofiber passive stress over a cardiac cycle. The new framework correctly predicted different forms of growth in response to two types of valvular diseases, namely aortic stenosis and mitral regurgitation. Furthermore, the model predicted that LV size and function are nearly restored (reversal of growth) when the disease-mimicking perturbation was removed in the simulations for each valvular disorder. In conclusion, the simulations suggest that time-averaged total stress along the fiber direction and time-averaged intracellular myofiber passive stress can be used to drive concentric and eccentric growth in simulations of valve disease.

Synergistic Biophysics and Machine Learning Modeling to Rapidly Predict Cardiac Growth Probability

Computational models that can predict growth and remodeling of the heart could have important clinical applications. However, the time it takes to calibrate and run current models while considering data uncertainty and variability makes them impractical for routine clinical use. This study aims to address this need by creating a computational framework to efficiently predict cardiac growth probability. We utilized a biophysics model to rapidly simulate cardiac growth following mitral valve regurgitation (MVR). Here we developed a two-tiered Bayesian History Matching approach augmented with Gaussian process emulators for efficient calibration of model parameters to align with growth outcomes within a 95% confidence interval. We first generated a synthetic data set to assess the accuracy of our framework, and the effect of changes in data uncertainty on growth predictions. We then calibrated our model to match baseline and chronic canine MVR data and used an independent data set to successfully validate the ability of our calibrated model to accurately predict cardiac growth probability. The combined biophysics and machine learning modeling framework we proposed in this study can be easily translated to predict patient-specific cardiac growth.

Computational Modelling Enabling In Silico Trials for Cardiac Physiologic Pacing

Conduction system pacing (CSP) has the potential to achieve physiological-paced activation by pacing the ventricular conduction system. Before CSP is adopted in standard clinical practice, large, randomised, and multi-centre trials are required to investigate CSP safety and efficacy compared to standard biventricular pacing (BVP). Furthermore, there are unanswered questions about pacing thresholds required to achieve optimal pacing delivery while preventing device battery draining, and about which patient groups are more likely to benefit from CSP rather than BVP. In silico studies have been increasingly used to investigate mechanisms underlying changes in cardiac function in response to pathologies and treatment. In the context of CSP, they have been used to improve our understanding of conduction system capture to optimise CSP delivery and battery life, and noninvasively compare different pacing methods on different patient groups. In this review, we discuss the in silico studies published to date investigating different aspects of CSP delivery.

Virtual pacing of a patient's digital twin to predict left ventricular reverse remodelling after cardiac resynchronization therapy

AIMS

Identifying heart failure (HF) patients who will benefit from cardiac resynchronization therapy (CRT) remains challenging. We evaluated whether virtual pacing in a digital twin (DT) of the patient's heart could be used to predict the degree of left ventricular (LV) reverse remodelling post-CRT.

METHODS AND RESULTS

Forty-five HF patients with wide QRS complex (≥130 ms) and reduced LV ejection fraction (≤35%) receiving CRT were retrospectively enrolled. Echocardiography was performed before (baseline) and 6 months after CRT implantation to obtain LV volumes and 18-segment longitudinal strain. A previously developed algorithm was used to generate 45 DTs by personalizing the CircAdapt model to each patient's baseline measurements. From each DT, baseline septal-to-lateral myocardial work difference (MWLW-S,DT) and maximum rate of LV systolic pressure rise (dP/dtmax,DT) were derived. Biventricular pacing was then simulated using patient-specific atrioventricular delay and lead location. Virtual pacing-induced changes ΔMWLW-S,DT and ΔdP/dtmax,DT were correlated with real-world LV end-systolic volume change at 6-month follow-up (ΔLVESV). The DT's baseline MWLW-S,DT and virtual pacing-induced ΔMWLW-S,DT were both significantly associated with the real patient's reverse remodelling ΔLVESV (r = -0.60, P < 0.001 and r = 0.62, P < 0.001, respectively), while correlation between ΔdP/dtmax,DT and ΔLVESV was considerably weaker (r = -0.34, P = 0.02).

CONCLUSION

Our results suggest that the reduction of septal-to-lateral work imbalance by virtual pacing in the DT can predict real-world post-CRT LV reverse remodelling. This DT approach could prove to be an additional tool in selecting HF patients for CRT and has the potential to provide valuable insights in optimization of CRT delivery.

A Computational Model of Ventricular Dimensions and Hemodynamics in Growing Infants

Previous computer models have successfully predicted cardiac growth and remodeling in adults with pathologies. However, applying these models to infants is complicated by the fact that they also undergo normal, somatic cardiac growth and remodeling. Therefore, we designed a computational model to predict ventricular dimensions and hemodynamics in healthy, growing infants by modifying an adult canine left ventricular growth model. The heart chambers were modeled as time-varying elastances coupled to a circuit model of the circulation. Circulation parameters were allometrically scaled and adjusted for maturation to simulate birth through 3 yrs of age. Ventricular growth was driven by perturbations in myocyte strain. The model successfully matched clinical measurements of pressures, ventricular and atrial volumes, and ventricular thicknesses within two standard deviations of multiple infant studies. To test the model, we input 10th and 90th percentile infant weights. Predicted volumes and thicknesses decreased and increased within normal ranges and pressures were unchanged. When we simulated coarctation of the aorta, systemic blood pressure, left ventricular thickness, and left ventricular volume all increased, following trends in clinical data. Our model enables a greater understanding of somatic and pathological growth in infants with congenital heart defects. Its flexibility and computational efficiency when compared to models employing more complex geometries allow for rapid analysis of pathological mechanisms affecting cardiac growth and hemodynamics.

Comparison of novel ventricular pacing strategies using an electro-mechanical simulation platform

AIMS

Focus of pacemaker therapy is shifting from right ventricular (RV) apex pacing (RVAP) and biventricular pacing (BiVP) to conduction system pacing. Direct comparison between the different pacing modalities and their consequences to cardiac pump function is difficult, due to the practical implications and confounding variables. Computational modelling and simulation provide the opportunity to compare electrical, mechanical, and haemodynamic consequences in the same virtual heart.

METHODS AND RESULTS

Using the same single cardiac geometry, electrical activation maps following the different pacing strategies were calculated using an Eikonal model on a three-dimensional geometry, which were then used as input for a lumped mechanical and haemodynamic model (CircAdapt). We then compared simulated strain, regional myocardial work, and haemodynamic function for each pacing strategy. Selective His-bundle pacing (HBP) best replicated physiological electrical activation and led to the most homogeneous mechanical behaviour. Selective left bundle branch (LBB) pacing led to good left ventricular (LV) function but significantly increased RV load. RV activation times were reduced in non-selective LBB pacing (nsLBBP), reducing RV load but increasing heterogeneity in LV contraction. LV septal pacing led to a slower LV and more heterogeneous LV activation than nsLBBP, while RV activation was similar. BiVP led to a synchronous LV-RV, but resulted in a heterogeneous contraction. RVAP led to the slowest and most heterogeneous contraction. Haemodynamic differences were small compared to differences in local wall behaviour.

CONCLUSION

Using a computational modelling framework, we investigated the mechanical and haemodynamic outcome of the prevailing pacing strategies in hearts with normal electrical and mechanical function. For this class of patients, nsLBBP was the best compromise between LV and RV function if HBP is not possible.

An in-silico analysis of experimental designs to study ventricular function: A focus on the right ventricle

In-vivo studies of pulmonary vascular disease and pulmonary hypertension (PH) have provided key insight into the progression of right ventricular (RV) dysfunction. Additional in-silico experiments using multiscale computational models have provided further details into biventricular mechanics and hemodynamic function in the presence of PH, yet few have assessed whether model parameters are practically identifiable prior to data collection. Moreover, none have used modeling to devise synergistic experimental designs. To address this knowledge gap, we conduct a practical identifiability analysis of a multiscale cardiovascular model across four simulated experimental designs. We determine a set of parameters using a combination of Morris screening and local sensitivity analysis, and test for practical identifiability using profile likelihood-based confidence intervals. We employ Markov chain Monte Carlo (MCMC) techniques to quantify parameter and model forecast uncertainty in the presence of noise corrupted data. Our results show that model calibration to only RV pressure suffers from practical identifiability issues and suffers from large forecast uncertainty in output space. In contrast, parameter and model forecast uncertainty is substantially reduced once additional left ventricular (LV) pressure and volume data is included. A comparison between single point systolic and diastolic LV data and continuous, time-dependent LV pressure-volume data reveals that at least some quantitative data from both ventricles should be included for future experimental studies.

Meshless Electrophysiological Modeling of Cardiac Resynchronization Therapy—Benchmark Analysis with Finite-Element Methods in Experimental Data

Computational models of cardiac electrophysiology are promising tools for reducing the rates of non-response patients suitable for cardiac resynchronization therapy (CRT) by optimizing electrode placement. The majority of computational models in the literature are mesh-based, primarily using the finite element method (FEM). The generation of patient-specific cardiac meshes has traditionally been a tedious task requiring manual intervention and hindering the modeling of a large number of cases. Meshless models can be a valid alternative due to their mesh quality independence. The organization of challenges such as the CRT-EPiggy19, providing unique experimental data as open access, enables benchmarking analysis of different cardiac computational modeling solutions with quantitative metrics. We present a benchmark analysis of a meshless-based method with finite-element methods for the prediction of cardiac electrical patterns in CRT, based on a subset of the CRT-EPiggy19 dataset. A data assimilation strategy was designed to personalize the most relevant parameters of the electrophysiological simulations and identify the optimal CRT lead configuration. The simulation results obtained with the meshless model were equivalent to FEM, with the most relevant aspect for accurate CRT predictions being the parameter personalization strategy (e.g., regional conduction velocity distribution, including the Purkinje system and CRT lead distribution).