Deep neural networks can classify ECGs with high accuracy when training data is abundant. Rare conditions like Brugada syndrome, an inherited arrhythmia syndrome predisposing to sudden death, pose challenges due to data scarcity hindering model training. We evaluated multiple machine learning (ML) approaches to optimise a Brugada ECG classification model using limited training data. The baseline model was trained on a dataset comprising 176 Brugada, 176 right bundle branch block (RBBB) and 352 normal ECGs from Zhongshan Hospital (Zhongshan-baseline dataset), framed as a binary classification task to distinguish Brugada from non-Brugada ECGs. A 25%-75% train-test split was used to exacerbate data scarcity. To enhance training, we incorporated three additional datasets: (i) a different, labelled ECG dataset from Zhongshan Hospital including normal and RBBB ECGs (Zhongshan-pretrain), (ii) an unlabelled ECG dataset from Hammersmith Hospital including Brugada and non-Brugada ECGs (Imperial), (iii) an open-access labelled ECG dataset (PTB-XL). Three strategies were tested: (1) supervised pretraining, (2) self-supervised pretraining with data augmentation, and (3) oversampling using SMOTE (synthetic minority oversampling technique). Each model was evaluated on the unseen internal test set and an external Brugada mimic dataset. The models were re-trained using an 80%-20% train-test split as a secondary analysis. The baseline model achieved 92.2% accuracy, F1-score 0.837, and area under the Receiver Operating Characteristic curve (AUC) 0.962. Supervised pretraining significantly improved performance when training data was scarce, with the best model pretrained on the Zhongshan-pretrain dataset boosting accuracy (+3.2%), F1-score (+0.071) and AUC + 0.019), with consistent cross-validation performance. Self-supervised pretraining produced smaller and more variable gains, although select models better mitigated against false positives on the Brugada mimic dataset. SMOTE oversampling showed inconsistent effects on performance. Incorporating pretraining and oversampling may facilitate the development of more accurate AI-ECG models for rare diseases when training data is limited but provides diminishing returns when adequate labelled data is available.

BACKGROUND:Hypertrophic cardiomyopathy (HCM) is associated with marked inter-patient heterogeneity in ventricular electrophysiology, contributing to arrhythmic risk that is insufficiently captured by current clinical methods. Electrocardiographic imaging (ECGI) provides high-density body surface potential (BSP) measurements but remains largely descriptive. Computational modelling offers a mechanistic framework to interpret BSP signals in terms of underlying tissue-level properties. METHODS AND FINDINGS:We developed a BSP-driven workflow to construct patient-specific electrophysiology (EP) models of HCM by integrating multimodal clinical imaging with Bayesian model calibration. Anatomically detailed torso-heart finite-element models were generated for 17 HCM patients using thoracic computed tomography (CT), cardiac magnetic resonance imaging (CMR), and 252-electrode BSP recordings. Ventricular depolarisation and repolarisation were simulated using a reaction-eikonal (RE) formulation coupled to a biophysically detailed ToR-ORd-dynCl ionic model. Emulator-based Bayesian history matching (HM) was used to personalise EP parameters, with staged calibration of QRS and T-wave morphology informed by targeted sensitivity analysis. The calibrated cohort reproduced clinical BSP morphology with Pearson correlation coefficient (PCC) [Formula: see text] for a median of 94.0% (IQR: 91.6 to 96.8%) of electrodes, achieving a median PCC of 0.89 (IQR: 0.80 to 0.94) across the full 252-electrode vest. Calibration substantially reduced uncertainty in the high-dimensional EP parameter space while yielding physiologically plausible conduction and repolarisation properties. Models calibrated exclusively to sinus rhythm robustly generalised to right-ventricular (RV) apical pacing without parameter retuning, reproducing clinically observed pacing-induced trends in depolarisation and repolarisation. Exploratory analysis revealed biologically consistent associations between inferred EP parameters and patient demographics. CONCLUSIONS:This study demonstrates that high-density BSP data can be used to functionally personalise mechanistic EP in HCM. The framework captures intrinsic patient-specific EP properties and generalises beyond the calibration condition, supporting its use for mechanistic investigation of arrhythmogenic substrate.

Dilated cardiomyopathy (DCM) and hypertrophic cardiomyopathy (HCM) are heart muscle diseases with largely opposing structural and functional phenotypes. Yet, both may lead to the same devastating outcomes of advanced heart failure and life-threatening arrhythmias. Using genome-wide association data from 9,365 DCM cases, 5,900 HCM cases, and over 1.2 million controls, we show that DCM and HCM are largely inversely associated across multiple genomic levels. Modeling both disorders as opposing genetic entities, in case-case GWAS approaches, we identify 100 loci (17 novel) underlying the cardiomyopathy spectrum. Several loci map to potential therapeutic targets (e.g., ADM, CACNA2D2), and polygenic risk scores derived from these data show strong discrimination between DCM and HCM patients in external datasets (AUC 0.78-0.84; AUPRC ~ 0.85). The pervasive opposing associations suggest that cardiomyocyte-directed therapies may often have opposite effects in DCM versus HCM. Nevertheless, a shared-effect analysis reveals a single locus - near the calcium-buffering gene CASQ2 - and also identifies a concordant genomic component associated with cardiometabolic health and extracardiac risk factors. By leveraging the shared and opposing genetic mechanisms of DCM and HCM, our work defines the genomic architecture of major cardiomyopathy subtypes and suggests new directions for therapeutics and precision medicine in heart failure.

BACKGROUND:Many patients with symptomatic obstructive hypertrophic cardiomyopathy (oHCM) have devices capable of right ventricular pacing (RVP). Although pacing can reduce left ventricular outflow tract gradient (LVOTg), it can also reduce cardiac output, so its net effect is variable. OBJECTIVES:We tested whether electromechanical optimization of the programmed atrio-ventricular delay (AVD) allows RVP to achieve a net benefit on symptoms. METHODS:EMORI-HCM (Electromechanically Optimized Right Ventricular Pacing in Obstructive Hypertrophic Cardiomyopathy) is a multicenter, blinded, randomized, crossover trial of AVD-optimized RVP in patients with symptomatic oHCM with resting or provoked gradient of at least 30 mm Hg. Patients with existing dual-chamber devices were randomized to either 3 months of continuous AVD-optimized RVP (intervention) followed by 3 months of backup-only RVP (control), or vice versa. AVD was optimized using a high-precision multiple-alternation protocol assessing acute change in beat-by-beat blood pressure while varying AVD. The primary outcome was symptoms measured by the Kansas City Cardiomyopathy Questionnaire Clinical Summary Score. Secondary outcomes include patient-reported daily symptom data collected using a dedicated smartphone application (ORBITA-app), dichotomous patient preference, EQ-5D, exercise capacity, and LVOTg. Patients were blinded to treatment allocation. Symptom assessments were self-administered. Outcome measures were recorded at baseline, crossover, and completion. Analysis was by Bayesian ordinal mixed modeling. RESULTS:Between October 2021 and October 2024, 117 screened patients met the inclusion criteria, of whom 60 were randomized. AVD-optimized RVP improved Kansas City Cardiomyopathy Questionnaire Clinical Summary Score (+4.5; 95% credible interval [CrI]: 1.3-8.1; probability of benefit [Prbenefit] = 0.997) and daily symptom scores (OR: 1.29; 95% CrI: 0.98-1.68; Prbenefit: 0.969) compared with backup-only pacing. AVD-optimized RVP improved exercise capacity (+1.0 mL/kg/min; 95% CrI: 0.1-2.0; Prbenefit: 0.984) and LVOTg (-7.3 mm Hg; 95% CrI: -13.5 to -1.1; Prbenefit: 0.010). It had no effect on B-type natriuretic peptide (Prbenefit: 0.893) and ejection fraction was preserved (Prbenefit: 0.409). CONCLUSIONS:In patients with oHCM, RVP delivered at electromechanically optimized AVD improves symptoms and exercise capacity. (Electromechanically Optimized Right Ventricular Pacing in Obstructive Hypertrophic Cardiomyopathy [EMORI-HCM], NCT05257772).

BACKGROUND:Patients with RBM20 cardiomyopathy present with an aggressive phenotype, associated with premature malignant arrhythmias, sudden cardiac death, and progressive heart failure (HF). This study aimed to investigate genotype-phenotype correlations, clinical outcomes, and causes of death in patients with RBM20-associated cardiomyopathy and review the current literature. METHODS:The cohort included patients with cardiomyopathy harboring pathogenic (P) or likely pathogenic (LP) RBM20 variants. For survival and regression analysis, a control group matched for sex, age, and presence of left ventricular dysfunction was included. Additionally, a comprehensive literature search was conducted. RESULTS:Sixty-two patients (45 % male, 42 ± 15 years at presentation) were included. We found 11 truncating variants. Patients with truncating variants diagnosed with HF were older compared to patients with missense variants (mean age 62 ± 9 vs. 45 ± 14; p = 0.01). Over a median follow-up duration of 5.0 [1.0-10.5] years, 21 (34 %) patients reached the composite endpoint, with 19 (31 %) patients experiencing malignant ventricular arrhythmia (VA) (mean age 45 ± 15 years, 63 % males). Males exhibited higher risk for the composite endpoint (log-rank p = 0.02), particularly for VA (log-rank p = 0.007). The literature review analyzed 34 studies comprising 678 patients (53 % male). In these studies, 123 (24 %) patients experienced a VA, 58 (12 %) underwent a heart transplant or were treated with LVAD, and 52 (11 %) died. CONCLUSION:This multicenter study highlights the severe phenotype associated with LP/P RBM20 variants, with a high incidence of VA, particularly in males. Additionally, this study presents 11 truncating variants mainly observed in older individuals.

BACKGROUND:Left ventricular outflow obstruction drives symptoms and outcomes in obstructive hypertrophic cardiomyopathy (oHCM). Right ventricular pacing (RVP) can desynchronize the left ventricle to relieve this and allows control of atrioventricular delay (AVD) but may impair ventricular function. We used high-precision assessment to quantify the hemodynamic and echocardiographic effects of RVP in oHCM. METHODS:Patients with oHCM and implanted dual-chamber pacing devices underwent continuous recording of ECG, beat-by-beat outflow tract continuous wave Doppler, and beat-by-beat, noninvasive finger-cuff blood pressure while pacing was alternated between atrium-only pacing and AV-sequential RVP at a range of AVDs at 5 bpm above resting heart rate and 100 bpm. Changes in systolic blood pressure (∆SBP) and left ventricular outflow tract gradient (∆LVOTg) were fitted to parabolas to produce reproducible, narrow confidence interval estimates of effects. RESULTS:Twenty two patients were recruited (60% male, mean resting LVOTg 53 mmHg). At just above resting heart rate (mean 75 bpm), RVP produced mean peak ∆SBP from AAI to DDD of 2.47 mmHg (95% confidence interval: 0.19-4.76, p = 0.04). The mean AVD for peak ∆SBP was 173.2 ms. Mean LVOTg reduction at this AVD was 8.31 mmHg (2.43-14.18, p < 0.001). Apart from the hemodynamically optimum AVD, no other AVDs produced statistically significant increases in SBP. At 100 bpm, greater increases in SBP and reductions in LVOTg were seen at hemodynamically optimal AVD. CONCLUSION:Multiple alternation assessment allows precise, reproducible, narrow confidence interval quantification of hemodynamic and echocardiographic pacing effects. RVP can reduce LVOTg while preserving or improving cardiac output, but AVD is a key modifier of this relationship.

Mitral valve prolapse (MVP) is the most common valvular heart disease and historically was considered benign in the absence of severe mitral regurgitation and normal left ventricular function. However, an emerging subset associated with sudden cardiac death that does not follow traditional risk stratification has been established termed the arrhythmic mitral valve prolapse (A-MVP) syndrome. This cohort provide a clinical challenge on identifying those at risk of lethal arrhythmias who may benefit from a primary prevention implantable cardiac defibrillator. We present 7 cases in which patients have survived cardiac arrest due to A-MVP, with the aim of further describing phenotypic characteristics that define this syndrome. We observed that after arrest, a high proportion of patients developed rapidly worsening mitral regurgitation. This observed link may provide insight into the underlying substrate of arrhythmogenesis.

Aims:Health systems face increasing waiting times for transcatheter aortic valve implantation (TAVI), incurring excess deaths and morbidity. To determine whether remote patient monitoring (RPM) using connected technologies can mitigate these risks by prioritizing patients awaiting TAVI, we aimed to measure the clinical safety and effectiveness of an RPM-based prioritization programme. Methods and results:Prospective observational cohort study of all patients awaiting TAVI at Imperial College Healthcare NHS Trust, London, UK, between 24th April 2023 and 15th November 2023. An RPM pathway was implemented for all patients accepted for TAVI. These patients responded to a weekly symptom questionnaire via web, smartphone RPM platform or telephone monitoring; with rule-based clinical escalation. The primary endpoint was the rate of adverse events (defined as emergency department presentation, unplanned hospitalization, or death), compared with a propensity score-matched (PSM) historical control group. Secondary endpoints included pathway performance characteristics for detection of deterioration. 200 patients met inclusion criteria. Despite growth of the waiting list, responsible for longer waiting times experienced by the RPM group [median 104 days (IQR 61.00-176.00) vs. 75 days (IQR 38.75-118.00)], there was no difference in rates of adverse events between RPM-patients and historical controls (Log rank P = 0.9). The RPM pathway had high sensitivity for prediction of waiting list death (100%). Patients deemed at high-risk of deterioration experienced shorter waiting times to treatment. Conclusion:RPM for patients awaiting TAVI is feasible and may mitigate the adverse effects of longer waiting times through accurate detection of deterioration and by informing prioritization decisions.

INTRODUCTION:Electrical activity in atrial fibrillation (AF) ranges from organized focal drivers to multiple wavelet re-entry. Understanding the effect of pulmonary vein isolation (PVI) on AF organization is clinically important, as it may optimize treatment strategies and outcomes. This study investigates the impact of PVI on AF organization and explores whether ventricular response regularity, measured from surface electrocardiograms (ECGs), reflects AF dynamics and electrophenotype. METHODS:Patients undergoing first-time PVI at Imperial College Healthcare NHS Trust between 2014 and 2022 were assessed pre- and post-PVI. AF organization was quantified using Shannon entropy (ShEn) and Sample entropy (SampEn) from coronary sinus (CS) electrograms. Ventricular response regularity was evaluated using surface ECG RR interval (RRI) variability and SampEn. RESULTS:PVI reduced ShEn and SampEn across all CS channels (e.g., SampEn CS 3-4: pre-ablation = 0.907 ± 0.512 vs. post-ablation = 0.790 ± 0.446, p < 0.001). Atrial ShEn and SampEn were correlated with ventricular response both pre- and post-ablation (e.g., correlations between atrial SampEn CS 3-4 and the following ventricular metrics: percentage of RRIs > 50 ms difference: r = 0.077, p = 0.008; normalized mean RRI difference: r = 0.144, p < 0.001; and normalized SampEn: r = 0.168, p < 0.001). CONCLUSION:The reduction in atrial ShEn and SampEn post-PVI indicates increased AF organization. The significant correlation between atrial entropy and ventricular variability suggests that AF organization affects ventricular response, assessed via surface ECG metrics. These findings highlight the potential of ECG-based measures as proxies for intracardiac AF organization.

BACKGROUND:Mitral valve prolapse (MVP) is the most common valvular heart disease and historically was considered benign in the absence of severe mitral regurgitation and normal left ventricular function. However, an emerging subset associated with sudden cardiac death that does not follow traditional risk stratification has been established termed the arrhythmic mitral valve prolapse (A-MVP) syndrome. This cohort provide a clinical challenge on identifying those at risk of lethal arrhythmias who may benefit from a primary prevention implantable cardiac defibrillator. CASE SUMMARY:We present 7 cases in which patients have survived cardiac arrest due to A-MVP, with the aim of further describing phenotypic characteristics that define this syndrome. DISCUSSION:We observed that after arrest, a high proportion of patients developed rapidly worsening mitral regurgitation. This observed link may provide insight into the underlying substrate of arrhythmogenesis.