Ahmed El-Medany

Wearable and portable electrocardiogram (ECG) devices are shifting cardiovascular monitoring beyond clinician-directed testing, driven by consumer-led self-monitoring, remote care, and preventive cardiology. Artificial intelligence (AI) may enhance the value of these devices by supporting automated interpretation, but clinical translation remains uneven. This narrative review synthesises current evidence on AI-enhanced ECG (AI-ECG) analysis in wearable and portable systems, with emphasis on the gap between technical performance and real-world implementation. Although many advances have been developed using conventional 12-lead ECG datasets, deployment in wearable and portable devices introduces additional challenges related to device-specific signal acquisition, preprocessing, domain shift, and user-led recording. Evidence is most mature for atrial fibrillation detection, while applications across conduction disease, other arrhythmias, structural phenotypes, and prognostic assessment remain earlier in development and are supported by fewer prospective, device-specific evaluations. Across the field, retrospective and dataset-based studies predominate, with limited evidence that improved detection translates into better patient outcomes or health-system efficiency. Future research should evaluate performance using the intended device, population, and clinical pathway, and determine whether AI-ECG changes management without increasing unnecessary investigation or workload. Establishing clinical value, rather than improving classification accuracy alone, will determine the role of wearable and portable AI-ECG in routine cardiovascular care.

Contrastive learning is a widely adopted self-supervised pretraining strategy, yet its dependence on cohort composition remains underexplored. We present Contrasting by Augmented Patient Electrocardiograms (CAPE) foundation model and pretrain on four cohorts (n = 5,203,269), from diverse populations across three continents (North America, South America, Asia). We systematically assess how cohort demographics, health status, and population diversity influence the downstream performance for prediction tasks also including two additional cohorts from another continent (Europe). We find that downstream performance depends on the distributional properties of the pretraining cohort, including demographics and health status. Moreover, while pretraining with a multi-centre, demographically diverse cohort improves in-distribution accuracy, it reduces out-of-distribution (OOD) generalisation of our contrastive approach by encoding cohort-specific artifacts. To address this, we propose the In-Distribution Batch (IDB) strategy, which preserves intra-cohort consistency during pretraining, discourages learning of spurious cohort-specific features, and instead promotes clinically meaningful variability within cohorts. This leads to improved out-of-distribution robustness, with gains of 9–40% in downstream label prediction performance. This work provides insights into pretraining strategies for more clinically deployable and generalisable foundation models.

Atrial cardiomyopathy (AtCM) is increasingly recognized as a substrate for atrial fibrillation (AF), yet its operationalization remains limited and largely marker-driven. Whether autonomic remodelling represents a longitudinally evolving domain within a multidomain framework of AtCM remains unclear. To determine whether autonomic remodelling, assessed through static and longitudinal heart rate variability (HRV) abnormalities, defines a distinct domain of AtCM and contributes to atrial disease burden and clinical risk. We studied 670 individuals aged 65 years without prior AF or major cardiovascular disease from the prospective PROOF cohort with 24-h Holter ECGs at baseline and 5 years. Heart rate variability metrics, premature atrial contraction (PAC) burden, and left atrial (LA) size were assessed. Incident AF and cardiovascular outcomes were adjudicated over a median follow-up of 12.1 years. Seventy-two participants (10.7%) developed AF. Static HRV abnormalities and adverse 5-year HRV trajectories were independently associated with subsequent AF. Autonomic abnormalities showed limited concordance with PAC burden and LA enlargement, supporting their role as a distinct AtCM domain. Increasing involvement of remodelling domains was associated with higher risks of AF and cardiovascular outcomes. Participants with ≥2 domains exhibited higher risks of AF (HR 2.43; 95% CI 1.72-3.45), major adverse cardiovascular events (HR 1.42), and all-cause mortality (HR 1.35). Atrial cardiomyopathy is a cumulative, multidomain disease process in which structural, electrical, and autonomic abnormalities define atrial disease burden. Longitudinal autonomic remodelling constitutes an independent and evolving axis within this framework, shifting the focus from isolated arrhythmia detection towards progressive characterization of atrial substrate.

Advances in implantable and wearable cardiac monitoring technologies have led to widespread detection of brief, often asymptomatic atrial high-rate episodes, frequently labelled as device-detected atrial fibrillation (AF). While detection has increased substantially, the clinical interpretation of these findings remains uncertain. Observational studies demonstrate associations between AF burden and stroke risk but reveal marked inter-individual heterogeneity and no consistent temporal threshold below which risk is eliminated. Recent randomised controlled trials show that anticoagulation guided solely by arrhythmia duration confers limited net clinical benefit, with modest reductions in ischaemic stroke offset by increased bleeding. These findings challenge the biological and clinical validity of rigid time-based thresholds for intervention. Increasing evidence suggests that AF may act primarily as a marker of underlying atrial disease rather than the sole mechanistic cause of thromboembolism. This article provides an evidence-informed perspective on the interpretation of device-detected AF in contemporary clinical practice and argues for a shift away from duration-based triggers toward a longitudinal, risk-adapted approach that integrates AF trajectory, atrial substrate, and clinical context. Emerging tools such as artificial intelligence-enhanced electrocardiography may help identify occult atrial pathology but must augment rather than replace clinical judgement. Proportionate, individualised care should supersede reflexive treatment strategies in the management of device-detected AF.

Mediastinal masses are occasionally involved in extrinsic compression of heart structures such as cardiac chambers, great vessels, and their branches. In rare cases, this can lead to haemodynamic compromise. Echocardiography is key to identifying and describing salient features that might explain a patient's symptoms, and contributes valuable prognostic information. Here, we report a case study of a young male patient who presented with cardiac symptoms and was referred to our outpatient department for a routine transthoracic echocardiogram revealing an unusual presentation of right ventricular outflow tract obstruction, pulmonary valve compression, and pulmonary stenosis secondary to a large extracardiac mass.

A rare case of pulmonary artery fibroelastoma that demonstrates the importance of multimodality imaging and serial scans in reducing diagnostic uncertainty.