Pulsed field ablation (PFA) is a promising new ablation modality for the treatment of atrial fibrillation (AF); however, the effect of fiber orientation on the ablation characteristics of PFA in AF treatment is still unclear, which is likely an essential factor in influencing the ablation characteristics. This study constructed an anatomy-based left atrium (LA) model incorporating fiber orientation and selected various electrical conductivity and ablation targets to investigate the effect of anisotropic electrical conductivity (AC), compared with isotropic electrical conductivity (IC), on the ablation characteristics of PFA in AF treatment. The results show that the percentage differences in the size of the surface ablation area between AC and IC are greater than 73.71%; the maximum difference in the size of the ablation isosurface between AC and IC at different locations in the atrial wall is 3.65 mm (X-axis), 3.65 mm (Z-axis), and 4.03 mm (X-axis), respectively; and the percentage differences in the size of the ablation volume are greater than 6.9%. Under the condition of the pulse, the amplitude is 1000 V, the total PFA duration is 1 s, and the pulse train interval is 198.4 ms; the differences in the temperature increase between AC and IC in LA are less than 2.46 °C. Hence, this study suggests that in further exploration of the computational study of PFA in AF treatment using the same or similar conditions as those used here (myocardial electrical conductivity, pulse parameters, and electric field intensity damage threshold), to obtain more accurate computational results, it is necessary to adopt AC rather than IC to investigate the size of the surface ablation area, the size of the ablation isosurface, or the size of the ablation volume generated by PFA in LA. Moreover, if only investigating the temperature increase generated by PFA in LA, adopting IC instead of AC for simplifying the model construction process is reasonable.

The non-thermal effects are considered one of the prominent advantages of pulsed field ablation (PFA). However, at higher PFA doses, the temperature rise in the tissue during PFA may exceed the thermal damage threshold, at which time intracardiac pulsatile blood flow plays a crucial role in suppressing this temperature rise. This study aims to compare the effect of heat dissipation of the different methods in simulating the pulsatile blood flow during PFA. This study first constructed an anatomy-based left atrium (LA) model and then applied the convective heat transfer (CHT) method and the computational fluid dynamics (CFD) method to the model, respectively, and the thermal convective coefficients used in the CHT method are 984 (W/m²*K) (blood-myocardium interface) and 4372 (W/m²*K) (blood–catheter interface), respectively. Then, it compared the effect of the above two methods on the maximum temperature of myocardium and blood, as well as the myocardial ablation volumes caused by irreversible electroporation (IRE) and hyperthermia under different PFA parameters. Compared with the CFD method, the CHT method underestimates the maximum temperature of myocardium and blood; the differences in the maximum temperature of myocardium and blood between the two methods at the end of the last pulse are significant (>1 °C), and the differences in the maximum temperature of blood at the end of the last pulse interval are significant (>1 °C) only at a pulse amplitude greater than 1000 V or pulse number greater than 10. Under the same pulse amplitude and different heat dissipation methods, the IRE ablation volumes are the same. Compared with the CFD method, the CHT method underestimates the hyperthermia ablation volume; the differences in the hyperthermia ablation volume are significant (>1 mm³) only at a pulse amplitude greater than 1000 V, a pulse interval of 250 ms, or a pulse number greater than 10. Additionally, the hyperthermia ablation isosurfaces are completely wrapped by the IRE ablation isosurfaces in the myocardium. Thus, during PFA, compared with the CFD method, the CHT method cannot accurately simulate the maximum myocardial temperature; however, except at the above PFA parameters, the CHT method can accurately simulate the maximum blood temperature and the myocardial ablation volume caused by IRE and hyperthermia. Additionally, within the range of the PFA parameters used in this study, the temperature rise during PFA may not lead to the appearance of additional hyperthermia ablation areas beyond the IRE ablation area in the myocardium.

Late gadolinium enhancement cardiac magnetic resonance imaging (LGE-CMR) is a valuable cardiovascular imaging technique. Segmentation of cardiac chambers from LGE-CMR is a fundamental step in electrophysiological modeling and cardiovascular disease diagnosis. Deep learning methods have demonstrated extremely promising performance. However, excellent performance often depended on a large amount of finely annotated data. The purpose of this manuscript was to develop a semi-supervised segmentation method to use unlabeled data to improve model performance. This manuscript proposed a semi-supervised network that integrates triple-consistency constraints (data-level, task-level, and feature-level) for cardiac chambers segmentation from LGE-CMR. Specifically, we designed a network that integrated segmentation and edge prediction tasks based on the mean teacher architecture. This addressed the problem of ignoring some challenging regions because of excluding low-confidence regions of previous research. We also applied a voxel-level contrastive learning strategy to achieve feature-level consistency, helping the model pay attention to the consistency between features overlooked in previous research. In terms of the Dice, Jaccard, Average Surface Distance (ASD), and 95% Hausdorff Distance (95HD) metrics, for the atrial segmentation dataset, the proposed method achieved scores of 88.34%, 79.30%, 7.92, and 2.02 when trained with 10% labeled data, and 90.70%, 83.09%, 6.41, and 1.72 when trained with 20% labeled data. For the ventricular segmentation task, the results were 87.22%, 77.95%, 2.27, and 0.61 with 10% labeled data, and 88.99%, 80.45%, 1.87, and 0.51 with 20% labeled data, respectively. Experiments demonstrated that our method outperforms previous semi-supervised methods, showing the potential of the proposed network for semi-supervised segmentation problems.

This study aims to investigate the mechanism of cardiac electrical activity through computer modeling and simulation. We optimized a comprehensive ventricular heart model and developed a whole ventricular electrophysiological model that incorporates myocardial stratification and a His-Purkinje network consistent with physiological structure. This was achieved by modifying the TT2 cell model parameters, utilizing tissue models, and incorporating a ventricular anatomy model. Subsequently, we conducted electrophysiological simulations of premature ventricular contractions (PVCs) and synthesized virtual electrocardiograms. The model effectively simulates the electrical activity of PVCs in various locations, including the right outflow tract interval, free wall, and left outflow tract. Furthermore, we compared the simulated electrical conduction mechanism of the human whole ventricle model with the synthetic virtual electrocardiogram. The synthesized electrocardiogram results exhibit a high degree of consistency with real clinical electrocardiogram morphology. The findings of this study provide insights into the electrocardiogram mechanism of PVCs.

Background: Cardiac contractility modulation (CCM) improves cardiac performance in heart failure, but its acute spatial effects on myocardial voltage and calcium-transient behavior remain incompletely characterized, particularly in the setting of early myocardial remodeling. Methods: In this exploratory, fixed-sequence study, isolated Langendorff-perfused hearts from Control and isoproterenol (ISO)-treated rats (n = 5 per group) underwent simultaneous voltage and calcium optical mapping during a 40-s protocol consisting of pre-stimulation (Base), active CCM stimulation (CCM-ON), and post-stimulation (CCM-OFF) phases. Global metrics were derived across the mapped ventricular surface and regional metrics from predefined regions proximal and distal to the stimulation electrodes. No time-matched sham-stimulation condition was included. Results: Relative optical action-potential (AP) and calcium-transient (CaT) amplitudes were approximately 5–6% higher during CCM-ON and CCM-OFF than during Base in both groups. Global AP duration at 90% repolarization (APD90) showed modest phase-dependent prolongation, whereas global CaT duration at 90% recovery (CTD90) showed modest shortening. Global CTD90 was longer overall in ISO-treated hearts, without a group-by-phase interaction. Three-way repeated-measures analyses identified phase-by-region interactions for CTD90 and AP–CaT delay (the interval from the 50% AP upstroke to the 50% CaT upstroke) and a group-by-region interaction for CTD90, with no significant phase-by-group-by-region interactions. CTD90 spatial dispersion and the absolute proximal–distal difference in AP–CaT delay showed phase effects, whereas mapped-surface APD90 dispersion did not change detectably. Conclusions: During the acute CCM sequence, CaT recovery and AP–CaT timing showed spatially nonuniform phase-associated patterns in Control hearts and ISO-treated hearts with features compatible with early remodeling, without a parallel detectable increase in mapped-surface APD90 dispersion. These exploratory, hypothesis-generating findings support a spatial dissociation between calcium-related temporal responses and mapped-surface APD90 dispersion under the present experimental conditions.