Background and Aims: Deep learning applied to electrocardiograms (ECG-AI) is an emerging approach for predicting atrial fibrillation or flutter (AF). This study introduces an ECG-AI model developed and tested at a tertiary cardiac centre, comparing its performance with clinical models and AF polygenic score (PGS). Methods: Electrocardiograms in sinus rhythm from the Montreal Heart Institute were analysed, excluding those from patients with pre-existing AF. The primary outcome was incident AF at 5 years. An ECG-AI model was developed by splitting patients into non-overlapping data sets: 70% for training, 10% for validation, and 20% for testing. The performance of ECG-AI, clinical models, and PGS was assessed in the test data set. The ECG-AI model was externally validated in the Medical Information Mart for Intensive Care-IV (MIMIC-IV) hospital data set. Results: A total of 669 782 ECGs from 145 323 patients were included. Mean age was 61 ± 15 years, and 58% were male. The primary outcome was observed in 15% of patients, and the ECG-AI model showed an area under the receiver operating characteristic (AUC-ROC) curve of .78. In time-to-event analysis including the first ECG, ECG-AI inference of high risk identified 26% of the population with a 4.3-fold increased risk of incident AF (95% confidence interval: 4.02–4.57). In a subgroup analysis of 2301 patients, ECG-AI outperformed CHARGE-AF (AUC-ROC = .62) and PGS (AUC-ROC = .59). Adding PGS and CHARGE-AF to ECG-AI improved goodness of fit (likelihood ratio test P < .001), with minimal changes to the AUC-ROC (.76–.77). In the external validation cohort (mean age 59 ± 18 years, 47% male, median follow-up 1.1 year), ECG-AI model performance remained consistent (AUC-ROC = .77). Conclusions: ECG-AI provides an accurate tool to predict new-onset AF in a tertiary cardiac centre, surpassing clinical and PGS.

Aims/hypothesis: Type 2 diabetes increases the risk of cardiovascular and renal complications, but early risk prediction could lead to timely intervention and better outcomes. Genetic information can be used to enable early detection of risk. Methods: We developed a multi-polygenic risk score (multiPRS) that combines ten weighted PRSs (10 wPRS) composed of 598 SNPs associated with main risk factors and outcomes of type 2 diabetes, derived from summary statistics data of genome-wide association studies. The 10 wPRS, first principal component of ethnicity, sex, age at onset and diabetes duration were included into one logistic regression model to predict micro- and macrovascular outcomes in 4098 participants in the ADVANCE study and 17,604 individuals with type 2 diabetes in the UK Biobank study. Results: The model showed a similar predictive performance for cardiovascular and renal complications in different cohorts. It identified the top 30% of ADVANCE participants with a mean of 3.1-fold increased risk of major micro- and macrovascular events (p = 6.3 × 10⁻²¹ and p = 9.6 × 10⁻³¹, respectively) and a 4.4-fold (p = 6.8 × 10⁻³³) higher risk of cardiovascular death. While in ADVANCE overall, combined intensive blood pressure and glucose control decreased cardiovascular death by 24%, the model identified a high-risk group in whom it decreased the mortality rate by 47%, and a low-risk group in whom it had no discernible effect. High-risk individuals had the greatest absolute risk reduction with a number needed to treat of 12 to prevent one cardiovascular death over 5 years. Conclusions/interpretation: This novel multiPRS model stratified individuals with type 2 diabetes according to risk of complications and helped to target earlier those who would receive greater benefit from intensive therapy. Graphical

Background and Aims: The 12-lead electrocardiogram (ECG) remains a cornerstone of cardiac diagnostics, yet existing artificial intelligence (AI) solutions for automated interpretation often lack generalizability, remain closed source, and are primarily trained using supervised learning (SL), which requires extensive labelled datasets and may limit adaptability across diverse clinical settings. Self-supervised learning (SSL) can potentially overcome these limitations by learning robust representations from unlabelled data. To address these challenges, this study developed and compared two open-source foundational ECG models: DeepECG-SL, a supervised multilabel ECG model, and DeepECG-SSL, a self-supervised model. Methods: Both models were trained on over 1 million ECGs using a standardized preprocessing pipeline and automated free-text extraction from ECG reports to predict 77 cardiac conditions. DeepECG-SSL leveraged unlabelled data through self-supervised contrastive learning and masked lead modelling before fine-tuning for downstream tasks, while DeepECG-SL was trained directly on labelled diagnostic data in an end-to-end fashion. Performance was evaluated across seven private, multilingual healthcare systems and four public ECG repositories, with assessment of fairness by age and sex, and investigation of privacy vulnerabilities as well as memory and compute requirements. Results: DeepECG-SSL achieved micro-averaged area under the receiver operating characteristic curves (AUROCs) across all 77 cardiac conditions for ECG interpretation of 0.990 [95% confidence interval (CI): 0.990, 0.990] on the internal dataset (MHI-ds), 0.981 (95% CI: 0.981, 0.981) on external public datasets (UKB, CLSA, MIMIC-IV and PTB), and 0.983 (95% CI: 0.983, 0.983) on external private datasets (UW, UCSF, JGH, NYP, MGH, CSH and CHUM), while DeepECG-SL demonstrated AUROCs of 0.992 (95% CI: 0.992, 0.992), 0.980 (95% CI: 0.980, 0.980), and 0.983 (95% CI: 0.983, 0.984), respectively. Fairness analyses revealed minimal disparities (true-positive rate and false-positive rate difference <0.1) across age and sex groups for both models. DeepECG-SSL demonstrated superior performance on limited-data digital biomarker tasks, with the largest improvements in long QT syndrome (LQTS) genotype classification (AUROC 0.931 vs 0.850, P = .026, n = 127 ECGs) and 5 year atrial fibrillation risk prediction (AUROC 0.742 vs 0.734, P < 0.001, n = 132 050 ECGs), while achieving superior performance in left ventricular ejection fraction ≤40% classification (AUROC 0.926 vs 0.917, P < 0.001, n = 25 252 ECGs) and comparable performance in LQTS detection (AUROC 0.767 vs 0.735, P = 0.117, n = 934 ECGs). Conclusions: This study establishes SSL as a promising paradigm for ECG analysis, particularly in settings with limited annotated data, enhancing accessibility, generalizability, and fairness in AI-driven cardiac diagnostics. By releasing model weights, preprocessing tools, and validation code, this work aims to support robust, data-efficient AI diagnostics across diverse clinical environments and questions.

Background: Natural selection eliminates detrimental and favors advantageous phenotypes. This process leaves characteristic signatures in underlying genomic segments that can be recognized through deviations in allelic or haplotypic frequency spectra. To provide an identifiable signature of recent positive selection that can be detected by comparison with the background distribution, we introduced a new way of looking at genomic polymorphisms: haplotype allelic classes. Results: The model combines segregating sites and haplotypic information in order to reveal useful data characteristics. We developed a summary statistic, Svd, to compare the distribution of the haplotypes carrying the selected allele with the distribution of the remaining ones. Coalescence simulations are used to study the distributions under standard population models assuming neutrality, demographic scenarios and selection models. To test, in practice, haplotype allelic class performance and the derived statistic in capturing deviation from neutrality due to positive selection, we analyzed haplotypic variation in detail in the locus of lactase persistence in the three HapMap Phase II populations. Conclusions: We showed that the Svd statistic is less sensitive than other tests to confounding factors such as demography or recombination. Our approach succeeds in identifying candidate loci, such as the lactase-persistence locus, as targets of strong positive selection and provides a new tool complementary to other tests to study natural selection in genomic data.