Friedman, Paul A.

Abstract Aims To assess the performance of an artificial intelligence-enabled electrocardiogram (AI-ECG) algorithm in identifying patients with moderate to severe aortic stenosis (AS) in an Asian cohort from a tertiary care centre. Methods and results We identified a randomly selected patients ≥60 years old who underwent echocardiography and ECG within in 31 days between 2012 and 2021 at the Samsung Medical Center in Korea. Patients with previous cardiac surgery, prosthetic valves, or pacemakers were excluded. The AI-ECG model, originally developed and validated by Mayo Clinic in the USA, was applied without fine-tuning. Performance metrics, including the area under the curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy, were calculated to compare AI-ECG predictions with TTE-confirmed AS status. Among 5425 patients, 1095 had moderate to severe AS, and 4330 age- and sex-matched patients without AS were included as controls. The AI-ECG model achieved an AUC of 0.85 (95% CI: 0.84–0.87) in detecting moderate to severe AS. Sensitivity, specificity, PPV, NPV, and accuracy were 0.83, 0.65, 0.37, 0.94, and 68.29%, respectively. The model's performance was consistent across various age and sex subgroups, with sensitivity increasing in older patients. Conclusion The AI-ECG model developed in the USA demonstrated comparable performance in detecting moderate to severe AS in an Asian cohort compared with its original validation population. These findings highlight the potential utility of AI-ECG as a non-invasive screening tool for AS across diverse patient populations.

ABSTRACT Background Transvenous lead removal (TLR) is a Class I indication for patients with cardiac implantable electronic device (CIED) infections. However, disparities in the utilization of TLR have not been fully explored, particularly concerning demographic and socioeconomic factors. Objectives This study aims to examine whether disparities exist among patients hospitalized for CIED infections with respect to the utilization of TLR, focusing on sex, race, and income characteristics. Methods We utilized the 2019 National Inpatient Sample (NIS) to assess patients hospitalized with CIED infections and identify those who underwent TLR. Multiple logistic regression was used to examine risk factors for TLR utilization, including age, sex, race/ethnicity, and socioeconomic characteristics. Results A total of 56 709 patients were hospitalized for CIED infection in the United States in 2019, of which 9867 (17.4%) underwent TLR. On an adjusted analysis, females and Black patients were found to have lower odds of undergoing TLR compared to males and White patients (OR: 0.62; 95% CI: 0.59–0.78; p < 0.01; and OR: 0.47; 95% CI: 0.39–0.57; p < 0.01, respectively). Hispanic patients also had reduced odds of TLR (OR: 0.73; 95% CI: 0.58–0.93; p = 0.01). No significant associations were observed between income or insurance status and the likelihood of receiving TLR after adjustment. Conclusions This study identified significant disparities in the utilization of TLR among patients hospitalized for CIED infections. Female, Black, and Hispanic patients were significantly less likely to receive TLR.

Abstract Aims Catheter ablation is commonly performed in patients with arrhythmogenic right ventricular cardiomyopathy (ARVC) for management of ventricular arrhythmias (VAs). The impact of catheter ablation on cardiac function is unknown. The aim is to assess the impact of catheter ablation for VAs on cardiac function in patients with ARVC. Methods and results This retrospective study included consecutive patients with ARVC who underwent catheter ablation for VAs. Cardiac and valvular functions were assessed on pre- and post-procedure cardiac imaging. The severities of tricuspid regurgitation, pulmonic regurgitation, right ventricular (RV) enlargement, and RV systolic dysfunction was graded as absent (=0), mild (=1), mild-moderate (=2), moderate (=3), moderate-severe (=4), or severe (=5). Thirty-four patients were included. Median age at the time of ablation was 49 [inter-quartile range (IQR), 23] years, and 10 (29%) patients were female. After a median follow-up of 1 (IQR, 0) day, 11 (34%) patients had worsening RV function, and 4 (12%) patients had worsening left ventricular ejection fraction (LVEF) &gt;5%. Five (15%) patients required inotropic support. After a median follow-up of 6.5 (IQR, 6.4) months, 20 (59%) patients had repeat transthoracic echocardiograms. Among patients with worsening RV function post-ablation, one-sixth (17%) had improvement in the RV function at follow-up. In contrast, among patients who had an acute drop in LVEF post-ablation, two-thirds (67%) had improvement at follow-up. Conclusion Right ventricular dysfunction following catheter ablation in patients with ARVC is common, affecting one-third of patients. Patients undergoing catheter ablation of VAs should be counselled on the potential negative impact of ablation on cardiac function.

Abstract Aims To develop a deep learning model that: (i) utilizes transthoracic echocardiography (TTE) clips to detect left ventricular (LV) enlargement without being provided information regarding a patient’s sex and body size; and (ii) can be accurately applied to clips acquired using either standard comprehensive TTE or handheld cardiac ultrasound (HCU). Methods and results Using retrospective TTE data (training: 8722 patients; internal validation: 468 patients), we developed a deep learning model that estimates a patient’s end-diastolic LV volume (indexed to body surface area and normalized across the sexes), and then thresholds this estimate to perform the following classifications: (1) normally sized LV vs. ≥ mild LV enlargement; (2) normal/mildly enlarged LV vs. ≥ moderate LV enlargement. For retrospective datasets, the model showed strong performance in TTE across three geographically distinct locations (Minnesota and Wisconsin: 1082 patients, AUC = 0.925 and 0.953 for classifications 1 and 2, respectively; Arizona: 1475 patients, AUC = 0.935 and 0.969; and Florida: 1481 patients, AUC = 0.934 and 0.970). Additionally, performance was strong for both TTE and HCU clips collected from a prospective cohort of 410 patients who underwent HCU immediately following TTE (TTE: AUC = 0.925 and 0.971; HCU: AUC = 0.874 and 0.902, for classifications 1 and 2, respectively). Conclusion An automated deep learning model applied to TTE or HCU images accurately categorizes LV volumes. These results lay a foundation for future work aimed at optimizing clinical outcomes for heart failure patients by enabling early detection of LV enlargement across various point-of-care settings.

Abstract Aims Artificial intelligence (AI) has the potential to transform cardiac electrophysiology (EP), particularly in arrhythmia detection, procedural optimization, and patient outcome prediction. However, a standardized approach to reporting and understanding AI-related research in EP is lacking. This scientific statement aims to develop and apply a checklist for AI-related research reporting in EP to enhance transparency, reproducibility, and understandability in the field. Methods and results An AI checklist specific to EP was developed with expert input from the writing group and voted on using a modified Delphi process, leading to the development of a 29-item checklist. The checklist was subsequently applied to assess reporting practices to identify areas where improvements could be made and provide an overview of the state of the art in AI-related EP research in three domains from May 2021 until May 2024: atrial fibrillation (AF) management, sudden cardiac death (SCD), and EP lab applications. The EHRA AI checklist was applied to 31 studies in AF management, 18 studies in SCD, and 6 studies in EP lab applications. Results differed between the different domains, but in no domain reporting of a specific item exceeded 55% of included papers. Key areas such as trial registration, participant details, data handling, and training performance were underreported (&lt;20%). The checklist application highlighted areas where reporting practices could be improved to promote clearer, more comprehensive AI research in EP. Conclusion The EHRA AI checklist provides a structured framework for reporting AI research in EP. Its use can improve understanding but also enhance the reproducibility and transparency of AI studies, fostering more robust and reliable integration of AI into clinical EP practice.

INTRODUCTION: Building on prior results, we hypothesized that an electrocardiogram (ECG)-enabled machine learning (ML) model could be used to detect advanced chronic liver disease (CLD). METHODS: A cohort with CLD and 12-lead ECGs was matched with controls from electronic health records. A ML model was trained as a binary classifier. RESULTS: There are 12,930 patients with CLD and 64,577 controls in the cohort. The model's discriminative ability to classify CLD showed an area under the receiver-operating characteristic curve 0.858 (95% confidence interval: 0.850–0.866), and at the chosen threshold, CLD ECGs had 12 times higher odds of being classified as CLD (diagnostic odds ratio 12.33, 95% confidence interval: 11.16–13.63). DISCUSSION: An ECG-enabled ML model affords great promise in identifying advanced CLD in low resource areas.

Abstract Background New postoperative atrial fibrillation (POAF) occurs in about 40% after cardiac surgery. Mineralocorticoid receptor antagonists (MRA) are known to reduce chronic atrial fibrillation (AF) development and burden. We examined the impact of preoperative MRA use on POAF and also examine the atrial cell type impacted by MRA treatment during cold cardiac preservation. Methods Retrospective study of 19,042 patients who underwent cardiac surgery at Mayo Clinic in Minnesota, and performed 1:3 propensity matching to obtain 298 patients on preoperative MRA matched to 894 who were not. We also separately matched patients using preoperative diuretics. Single-nuclei RNA sequencing (snRNA-seq) examined MRA’s effects on different atrial cell types in canrenone (water soluble MRA) treated human donor hearts undergoing cold preservation followed by ex-vivo reperfusion and compared gene expression to the atria of patients with AF. Results Propensity matched preoperative MRA group had less new onset POAF (19.8% vs 31.5%, P<0.001). To account for the possibility that preoperative diuretic use and volume reduction may impact POAF, we propensity matched 298 preop diuretic users that included MRA use to another 894 patients who used a non-MRA diuretic preoperatively. Those who used preoperative MRA similarly had a lower incidence of POAF (19.8% vs 33.2%, P<0.001). No survival difference was present between the propensity matched groups that used preoperative diuretics (P=0.079). Preoperative MRA use also reduced the development of paroxysmal and chronic AF at 6 years of follow up. From our snRNA-seq data, we identified a subpopulation of atrial cardiomyocytes (CM2) that had high MR expression where canrenone suppressed the increase in MR target gene expression associated with cold preservation-reperfusion. These MR targets were conversely elevated in patients with chronic AF. Canrenone also suppressed other cardiac preservation associated genes that show elevated expression in atrial macrophages and pericytes from chronic AF atria. Conclusions Our studies show that preoperative MRA use is associated with 40% reduction in POAF as well as lowering long standing AF development by about 41%. Our cold cadiac preservation-reperfusion model showed that canrenone reduced expression of MR target genes associated with chronic AF, particular in cardiomyocytes with important roles in electrical conduction. Clinical Perspective What is New? This study shows that preoperative use of mineralocorticoid receptors antagonists (MRA) is associated with a reduced incidence of new onset perioperative atrial fibrillation after cardiac surgery utilizing cardiopulmonary bypass. We show that preoperative MRA use is associated with a lower incidence of developing more chronic paroxymal or sustained atrial fibrillation. Addition of canrenone, a clinically utilized water soluble MRA, to cardioplegia solution used during cardiac preservation can attenuate atrial inflammatory reponses and reduce signaling through molecular pathways that promote atrial fibrillation. What are the clinical implications? Perioperative use of MRAs may be considered to reduce early postoperative atrial fibrillation as well as lowering the risk of developing more chronic atrial arrhythmias. These findings support pursuing a clinical trial to determine the impact of MRA use on atrial arrhythmias following cardiac surgery in the setting of cardiopulmonary bypass with cold cardiac preservation.

Background We tested whether biologic age, as estimated by deficits, functional impairments, or Age‐Gap or their combination, provide improved estimation of cause‐specific death as compared with chronological age. Methods Cardiovascular and noncardiovascular deficits, functional impairments, and Age‐Gap were prospectively collected in 535 patients aged ≥55 years undergoing percutaneous coronary interventions between August 1, 2014, and March 31, 2018. Age‐Gap was calculated as the difference between chronological age and age estimated by artificial intelligence ECG using a convolutional neural network. The full biological age model included deficits, functional impairments, and Age‐Gap >2 SD. A multivariable reduced model with the least number of variables was also created to provide a comparable C index to the full model. Results The average chronological age was 72.1±9.5 years, and there were 68% of men. During a median follow‐up of 2.61 years, 124 (23%) patients died. There was a modest correlation between Age‐Gap and biological age ( r =0.28 [95% CI, 0.20–0.35]; P <0.001). When modeled with chronologic age as a covariate, Age‐Gap predicted all‐cause (hazard ratio [HR], 1.07 [95% CI, 1.04–1.10]; P <0.001) and cardiovascular (HR, 1.07 [95% CI, 1.04–1.11]; P <0.001) mortality. As compared with chronological age, the full biological age model noted significant improvement in the prediction of long‐term overall (95% CI, 0.65–0.78), cardiovascular (95% CI, 0.69–0.77), and noncardiovascular (95% CI, 0.55–0.86) mortality. In the reduced models, most prognostic information for noncardiovascular mortality (C index: 0.79) was obtained by subjective difficulty in performing tasks, whereas the deficit‐based estimation predicted cardiovascular mortality (C index: 0.72). Conclusions Estimated biological age from deficits and functional impairments was superior to chronological age in predicting long‐term cause‐specific mortality following percutaneous coronary interventions.