Daily Cardiology Research Analysis
Analyzed 31 papers and selected 3 impactful papers.
Summary
Three high-impact cardiology papers stand out today: a prospective CMR study shows extracellular volume fraction (ECV%) independently predicts 10-year mortality after aortic valve replacement in severe aortic stenosis; a multisite external validation confirms an AI-enabled ECG can accurately detect low ejection fraction and function as a rule-out tool; and a wearable-derived AI-ECG “age” gap correlates with atrial fibrillation presence and burden, supporting its role as a digital biomarker.
Research Themes
- AI-enabled cardiovascular diagnostics and screening
- CMR-derived myocardial fibrosis biomarkers for risk stratification
- Wearable digital biomarkers for arrhythmia risk and burden
Selected Articles
1. Extracellular volume fraction associates with long-term outcome in patients with severe symptomatic aortic stenosis: 10-year outcomes of the RELIEF-AS Study.
In a single-center prospective cohort of 168 patients with severe symptomatic aortic stenosis undergoing AVR, higher baseline ECV% by CMR T1 mapping independently predicted 10-year all-cause mortality alongside age and atrial fibrillation. Patients who died had higher ECV% and LGE; the parsimonious multivariable model favored age, AF, and ECV% as key predictors.
Impact: This is the longest follow-up CMR T1 mapping study in severe AS showing diffuse fibrosis (ECV%) remains an independent predictor of mortality post-AVR, directly informing risk stratification beyond traditional surgical scores.
Clinical Implications: Integrating ECV% into pre-AVR assessment could refine prognostic stratification, identify high-risk patients for closer follow-up, and guide timing or adjunctive therapies aimed at fibrosis.
Key Findings
- Higher baseline ECV% was associated with increased 10-year all-cause mortality (29.9% vs 27.6%, p=0.014).
- In multivariable Cox analysis, age, atrial fibrillation, and ECV% were independent predictors of mortality.
- Late gadolinium enhancement and ECV% were higher in those who died, but a parsimonious model favored age, AF, and ECV% (AIC-based selection).
Methodological Strengths
- Prospective observational design with long-term (median ~10 years) follow-up.
- Standardized multiparametric assessment including CMR T1 mapping for quantitative ECV% and national database linkage for mortality.
Limitations
- Single-center study with a moderate sample size (n=168) limits generalizability.
- Observational design cannot establish causality; potential selection and residual confounding.
Future Directions: Validate ECV%-based risk models across multicenter cohorts; test whether fibrosis-modifying therapies or altered timing of AVR improve outcomes in high-ECV% patients.
AIMS: Diffuse fibrosis is central to the pathophysiology of aortic stenosis (AS), can be assessed using cardiovascular magnetic resonance (CMR) with extracellular volume fraction (ECV%), and associates with mortality. The relevance of this signal to long-term prognosis remains unclear. We aim to assess predictors of long-term mortality with focus on diffuse fibrosis. METHODS AND RESULTS: Single-centre prospective observational cohort study of patients with severe, symptomatic AS undergoing AVR. Patients were assessed using echocardiography, high-sensitivity cardiac troponin T (hs-cTnT), N-terminal pro-B type natriuretic peptide (NT-proBNP) and CMR including T1 mapping for ECV% quantification. All-cause mortality was identified using the NHS National Spine Database. Univariable and multivariable Cox regression models were fitted to assess all-cause mortality associations. 168 patients (age 72 [65-77] years, 55% male) underwent CMR. Over a follow-up period of 9.7 (6.8-10.9) years, 76 deaths occurred. Patients who died had higher ECV% (29.9% vs 27.6%, p=0.014) and greater LGE (3.9% vs 2.0%, p=0.013). Univariable predictors of mortality were age, atrial fibrillation (AF), left atrial area, left atrial volume, total cholesterol, triglycerides, HDL:LDL ratio, non-bicuspid aortic valve, hs-cTnT, NT-proBNP, EuroSCORE II and ECV%. On multivariable regression, age, AF and ECV% remained significant predictors of mortality, independently of sex. AIC indicated that the model with four covariates was preferable to the one also including EuroSCORE II and coronary artery disease, and this result was confirmed by a likelihood ratio test (p=0.387). CONCLUSIONS: In the longest follow-up cohort of T1 mapping in severe AS, we demonstrate diffuse fibrosis remains an independent predictor of long-term mortality. Integration of ECV% in baseline risk stratification should be explored further in patients with AS undergoing AVR.
2. Multisite, External Validation of an AI-Enabled ECG Algorithm for Detection of Low Ejection Fraction.
Across 13,960 patients from four U.S. sites with ECG and echocardiography within 30 days, an FDA-style ECG-AI device achieved AUROC 0.92, sensitivity 84.5%, specificity 83.6%, and NPV 98.4% for detecting low EF. The algorithm yielded test negatives in 78% of cases, supporting use as a rule-out tool to reduce unnecessary echocardiography when clinical suspicion is low.
Impact: This large, multisite external validation provides robust, practice-ready evidence for AI-ECG as a scalable screening tool to detect low EF, with immediate potential to optimize echocardiography utilization.
Clinical Implications: AI-ECG could serve as a high-NPV rule-out test in routine care, triaging patients for echocardiography and enabling earlier identification of asymptomatic LV dysfunction.
Key Findings
- AUROC 0.92 (95% CI 0.91–0.93) with sensitivity 84.5% and specificity 83.6% for detecting low EF.
- Negative predictive value 98.4% and positive predictive value 30.5% with a 7.9% prevalence of low EF.
- 78% of patients tested negative, supporting AI-ECG as a rule-out strategy to defer echocardiography when clinical findings are absent.
Methodological Strengths
- Geographically diverse, multisite external validation with a large sample size.
- Standardized data extraction and predefined pairing of ECG with echocardiography within 30 days.
Limitations
- Retrospective design without patient-level clinical outcomes beyond diagnostic performance.
- Generalizability to non-U.S. settings, varied ECG devices, and demographic subgroups requires further validation.
Future Directions: Prospective impact studies to assess workflow integration, cost-effectiveness, and patient outcomes; evaluation across broader care settings and devices; calibration for specific subpopulations.
BACKGROUND: Low left ventricular ejection fraction (LEF) can progress undiagnosed. Artificial intelligence-based electrocardiogram (ECG-AI) screening may provide a scalable means to detect LEF. OBJECTIVES: The purpose of this study was to validate a complete ECG-AI software as a medical device for LEF detection. METHODS: Four geographically diverse sites in the United States identified patients with both ECGs and transthoracic echocardiograms performed within 30 days of each other in clinical practice. Data were electronically extracted to specific guidelines and transmitted to the coordinating center for analysis. RESULTS: Records of 16,000 subjects were extracted, resulting in an evaluable set of 13,960 subjects (mean age 66 years; 52% male). The device demonstrated excellent discrimination (AUROC: 0.92 [95% CI: 0.91-0.93]) and was 84.5% (95% CI: 82.2%-86.6%) sensitive and 83.6% (95% CI: 82.9%-84.2%) specific for LEF. The overall prevalence of LEF in the study data set was 7.9%, with LEF among 1.6% of the ECG-AI negative and 30.5% of ECG-AI positive subjects, contributing to positive and negative predictive values of 30.5% (95% CI: 28.8%-32.1%) and 98.4% (95% CI: 98.2%-98.7%), respectively. CONCLUSIONS: External validation studies such as this one provide a rigorous framework to validate an algorithm's performance. This study demonstrated the algorithm's strong diagnostic accuracy over a geographically diverse, independent set of patients. In this generally unselected population, the algorithm produced a test negative result in 78% of the cases, suggesting potential utility as a rule-out strategy to defer echocardiography when other clinical findings are absent.
3. Wearable device derived electrocardiographic age and its association with atrial fibrillation.
Using 1 million hospital ECGs transformed to synthetic single-lead signals for training, a wearable AI model estimated ECG age with MAE ~10–12 years. In two registered, independent wearable cohorts, each 1-year ECG-age gap increased AF odds (OR 1.03) and was associated with 0.8 percentage point higher AF burden, supporting ECG-age as a proactive digital biomarker.
Impact: This study operationalizes a wearable AI biomarker by bridging hospital-scale training with real-world self-monitoring validation, highlighting a scalable pathway for AF risk and burden surveillance.
Clinical Implications: Wearable AI-ECG age could enrich AF screening strategies by flagging individuals with higher biological ECG-age for intensified rhythm monitoring, potentially enabling earlier AF detection and management.
Key Findings
- AI model trained on 1,000,000 hospital 12-lead ECGs converted to synthetic single-lead signals via CycleGAN.
- Validation in two wearable cohorts yielded mean absolute errors of 10.01 and 11.88 years for ECG age estimation.
- Each 1-year ECG-age gap associated with higher AF odds (OR 1.03) and 0.8 percentage point increase in AF burden.
Methodological Strengths
- Bridging large-scale clinical ECG training with independent, registered wearable validation cohorts.
- Use of generative modeling (CycleGAN) and residual networks to adapt 12-lead data to single-lead wearable constraints.
Limitations
- Effect sizes per 1-year age gap are modest; clinical thresholds and actionability require prospective evaluation.
- Validation cohort sizes are not specified in the abstract; generalizability across devices and populations needs further study.
Future Directions: Define actionable ECG-age thresholds in prospective studies, integrate with clinical risk scores, and evaluate impact on AF detection, burden reduction, and outcomes.
Artificial intelligence (AI)-derived electrocardiographic (ECG) age is a promising marker of atrial fibrillation (AF) risk. We developed PROPHECG-Age Single-an AI model estimating ECG age from wearable single-lead ECGs-and examined whether the ECG-age gap (predicted minus chronological age) is associated with AF presence and burden in real-world self-monitoring context. One million 12-lead ECGs from a hospital were converted to synthetic single-lead signals via Cycle-Consistent Generative Adversarial Network and used to train a residual network-based model. Validation in two independent wearable cohorts (S-Patch [ClinicalTrials.gov: NCT05119725, registered November 2021]; Memo Patch [ClinicalTrials.gov: NCT05355948, registered May 2022]) showed mean absolute errors of 10.01 and 11.88 years, respectively. The pooled association with AF presence was significant (odds ratio 1.03 per 1-year gap), and for AF burden, each 1-year gap increase corresponded to a 0.8 percentage point rise. These findings support wearable-based AI-ECG age as a potential digital biomarker for proactive cardiovascular monitoring.