Search for: All records

Cardiac myocytes synchronize through electrical signaling to contract heart muscles, facilitated by gap junctions (GJs) located in the intercalated disc (ID). GJs provide low-resistance pathways for electrical impulse propagation between myocytes, considered the primary mechanism for electrical communication in the heart. However, research indicates that conduction can persist without GJs. Ephaptic coupling (EpC), which depends on electrical fields in the narrow ID between adjacent myocytes, offers an alternative mechanism for cardiac conduction when GJs are impaired. Research suggests that EpC can enhance conduction velocity (CV) and reduce the likelihood of conduction block (CB), particularly when GJs are impaired, demonstrating the anti-arrhythmic potential of EpC. Reduced GJ communication increases the susceptibility of heart to arrhythmias due to ectopic or triggered activity, highlighting the pro-arrhythmic effect of GJ uncoupling. However, the interplay between GJs and EpC, and their roles in the initiation, dynamics, and termination of arrhythmias, remain unclear. Reentry, characterized by a loop of electrical activity, is a common mechanism underlying arrhythmogenesis in the heart. This study aims to explore the interplay between EpC and GJs on reentry initiation and its underlying dynamics. Specifically, we employed a two-dimensional (2D) discrete bidomain model that integrates EpC to simulate ephaptic conduction during reentry. We quantitatively assessed the outcomes of reentry initiation and the resulting dynamics across different levels of EpC, GJs, and initial perturbations. The results show that sufficiently strong EpC (i.e., sufficiently narrow clefts) tends to suppress reentry initiation, resulting in absent or non-sustained reentrant activity, while also introducing transient instability and heterogeneity into the cardiac dynamics. In contrast, relatively weak EpC (wide clefts) support sustained reentry with a stable rotor. Furthermore, we found that sufficiently strong EpC can lower the maximal dominant frequency observed during reentrant activity. Overall, this suggests that strong EpC exerts an anti-arrhythmic effect. 

Circadian clocks regulate many aspects of human physiology, including cardiovascular function and drug metabolism. Administering drugs at optimal times of the day may enhance effectiveness and reduce side effects. Certain cardiac antiarrhythmic drugs have been withdrawn from the market due to unexpected proarrhythmic effects such as fatal Torsade de Pointes (TdP) ventricular tachycardia. The Comprehensive in vitro Proarrhythmia Assay (CiPA) is a recent global initiative to create guidelines for the assessment of drug-induced arrhythmias that recommends a central role for computational modeling of ion channels andin silicoevaluation of compounds for TdP risk. We simulated circadian regulation of cardiac excitability and explored how dosing time of day affects TdP risk for 11 drugs previously classified into risk categories by CiPA. The model predicts that a high-risk drug taken at the most optimal time of day may actually be safer than a low-risk drug taken at the least optimal time of day. Based on these proof-of-concept results, we advocate for the incorporation of circadian clock modeling into the CiPA paradigm for assessing drug-induced TdP risk. Since cardiotoxicity is the leading cause of drug discontinuation, modeling cardiac-related chronopharmacology has significant potential to improve therapeutic outcomes. 

Generating realistic audio for human actions is critical for applications such as film sound effects and virtual reality games. Existing methods assume complete correspondence between video and audio during training, but in real-world settings, many sounds occur off-screen or weakly correspond to visuals, leading to uncontrolled ambient sounds or hallucinations at test time. This paper introduces AV-LDM, a novel ambient-aware audio generation model that disentangles foreground action sounds from ambient background noise in in-the-wild training videos. The approach leverages a retrieval-augmented generation framework to synthesize audio that aligns both semantically and temporally with the visual input. Trained and evaluated on Ego4D and EPIC-KITCHENS datasets, along with the newly introduced Ego4D-Sounds dataset (1.2M curated clips with action-audio correspondence), the model outperforms prior methods, enables controllable ambient sound generation, and shows promise for generalization to synthetic video game clips. This work is the first to emphasize faithful video-to-audio generation focused on observed visual content despite noisy, uncurated training data.