Daily Cardiology Research Analysis
AI-enabled cardiology made major strides: an AI model using single-view point-of-care ultrasound accurately screened for hypertrophic and amyloid cardiomyopathies years before diagnosis, while a machine-learning model predicted 1-year stroke/death after TCAR with excellent discrimination. Translationally, inhibiting ACLY with bempedoic acid prevented abdominal aortic aneurysm formation in mice, suggesting a repurposable pathway for a disease lacking medical therapy.
Summary
AI-enabled cardiology made major strides: an AI model using single-view point-of-care ultrasound accurately screened for hypertrophic and amyloid cardiomyopathies years before diagnosis, while a machine-learning model predicted 1-year stroke/death after TCAR with excellent discrimination. Translationally, inhibiting ACLY with bempedoic acid prevented abdominal aortic aneurysm formation in mice, suggesting a repurposable pathway for a disease lacking medical therapy.
Research Themes
- AI-enabled cardiac screening and prognostication
- Peri-procedural risk stratification using machine learning
- Drug repurposing targeting inflammatory-metabolic pathways in vascular disease
Selected Articles
1. Artificial intelligence-guided detection of under-recognised cardiomyopathies on point-of-care cardiac ultrasonography: a multicentre study.
A video-based, single-view-capable AI model applied to cardiac POCUS discriminated hypertrophic cardiomyopathy and transthyretin amyloid cardiomyopathy with AUCs ~0.90–0.97 across two health systems and flagged cases a median of ~2 years before clinical diagnosis. In patients without known cardiomyopathy, higher AI scores independently predicted mortality over a median 2.8 years, supporting opportunistic screening with simple POCUS acquisitions.
Impact: This work operationalizes scalable AI screening for underdiagnosed cardiomyopathies using real-world POCUS, potentially enabling earlier detection and risk stratification without comprehensive echocardiography.
Clinical Implications: Emergency and community settings could deploy AI-assisted POCUS to triage patients for confirmatory imaging, genetics or biopsy, initiate earlier disease-modifying therapy (e.g., ATTR therapies), and identify high-risk individuals for closer follow-up.
Key Findings
- Single-view AI on POCUS discriminated HCM and ATTR cardiomyopathy with AUCs ~0.90–0.97 across independent health systems.
- AI-positive screens preceded clinical diagnosis by a median 2.1 years (HCM) and 1.9 years (ATTR).
- Among 25,261 individuals without known cardiomyopathy, top-quintile AI scores for HCM and ATTR associated with higher adjusted mortality (HR 1.17 and 1.39, respectively).
- Model used a multi-label video CNN with view-quality–weighted loss to adapt to POCUS variability.
Methodological Strengths
- Very large development corpus (290,245 echo videos) with external validation across systems
- Customized loss and single-view protocol validated for real-world POCUS variability
Limitations
- Retrospective design with potential selection bias and lack of prospective clinical impact trial
- Generalizability beyond two US health systems and to handheld devices requires testing
Future Directions: Prospective implementation trials to test clinical pathways, confirm diagnostic yield, outcomes impact, equity/fairness, and cost-effectiveness across diverse care settings.
BACKGROUND: Point-of-care ultrasonography (POCUS) enables cardiac imaging at the bedside and in communities but is limited by abbreviated protocols and variation in quality. We aimed to develop and test artificial intelligence (AI) models to screen for under-diagnosed cardiomyopathies from cardiac POCUS. METHODS: In a development set of 290 245 transthoracic echocardiographic videos across the Yale-New-Haven Health System (YNHHS), we used augmentation approaches, and a customised loss function weighted for view quality to derive a POCUS-adapted, multi-label, video-based convolutional neural network that discriminates hypertrophic cardiomyopathy and transthyretin amyloid cardiomyopathy from controls without known disease. We evaluated the model across independent, internal, and external, retrospective cohorts of individuals undergoing cardiac POCUS across YNHHS and the Mount Sinai Health System (MSHS) emergency departments (between 2012 and 2024) to prioritise key views and validate the diagnostic and prognostic performance of single-view screening protocols. FINDINGS: Between Nov 1, 2023, and March 28, 2024, we identified 33 127 patients ...
2. Using machine learning to predict outcomes following transcarotid artery revascularization.
Across 38,325 TCAR procedures, an XGBoost model using 115 perioperative features predicted 1-year stroke/death with AUROC 0.91 pre-operatively and up to 0.94 post-operatively, outperforming logistic regression by a wide margin. The tool supports perioperative risk mitigation and individualized decision-making for a high-risk vascular population.
Impact: Delivers a validated, high-discrimination ML risk model at scale in a real-world registry for a technically demanding procedure where outcome prediction tools are scarce.
Clinical Implications: Preoperative and intra/postoperative ML scores could guide selection, optimization, resource allocation (ICU monitoring, neuroprotection), and counseling; integration into vascular workflow may reduce adverse events.
Key Findings
- In 38,325 TCAR cases, 7.0% had 1-year stroke or death; XGBoost predicted this with AUROC 0.91 using preoperative data.
- Performance improved with intra- and postoperative features (AUROCs 0.92 and 0.94), substantially outperforming logistic regression (AUROC 0.68).
- Model trained with tenfold cross-validation using 115 features spanning pre-, intra-, and postoperative phases.
Methodological Strengths
- Very large national registry with comprehensive perioperative features
- Systematic comparison to conventional modeling and staged (pre/intra/post) evaluation
Limitations
- Retrospective registry design without external validation beyond VQI; risk of dataset shift
- Model interpretability and clinical impact not prospectively tested
Future Directions: Prospective external validation, integration into clinical workflows with impact evaluation, calibration drift monitoring, and explainability to support clinician adoption.
Transcarotid artery revascularization (TCAR) is a relatively new and technically challenging procedure that carries a non-negligible risk of complications. Risk prediction tools may help guide clinical decision-making but remain limited. We developed machine learning (ML) algorithms that predict 1-year outcomes following TCAR. The Vascular Quality Initiative (VQI) database was used to identify patients who underwent TCAR between 2016 and 2023. We identified 115 features from the index hospitalization (82 pre-operative [demographic/clinical], 14 intra-operative [procedural], and 19 post-operative [in-hospital course/complications]). The primary outcome was 1-year post-procedural stroke or death. The data was divided into training (70%) and test (30%) sets. Six ML models were trained using pre-operative features with tenfold cross-validation. Overall, 38,325 patients were included (mean age 73.1 [SD 9.0] years, 14,248 [37.2%] female) and 2,672 (7.0%) developed 1-year stroke or death. The best pre-operative prediction model was XGBoost, achieving an AUROC of 0.91 (95% CI 0.90-0.92). In comparison, logistic regression had an AUROC of 0.68 (95% CI 0.66-0.70). The XGBoost model maintained excellent performance at the intra- and post-operative stages, with AUROC's (95% CI's) of 0.92 (0.91-0.93) and 0.94 (0.93-0.95), respectively. Our ML algorithm has potential for important utility in guiding peri-operative risk-mitigation strategies to prevent adverse outcomes following TCAR.
3. Inhibition of ATP-citrate lyase by bempedoic acid protects against abdominal aortic aneurysm formation in mice.
Active ACLY was upregulated in human AAA inflammatory infiltrates and in AngII-induced aneurysms in ApoE−/− mice. Pharmacologic inhibition of ACLY with the clinically available agent bempedoic acid protected against AAA formation in mice, implicating immunometabolic ACLY signaling as a therapeutic target for a disease currently lacking medical therapy.
Impact: Suggests a repurposable, clinically available ACLY inhibitor to prevent or slow AAA, a high-burden condition without proven medical therapy.
Clinical Implications: If translated, bempedoic acid or ACLY pathway modulation could offer a pharmacologic option to reduce AAA growth or incidence, complementing surveillance and surgical/endovascular repair.
Key Findings
- Active (phosphorylated) ACLY is increased in human AAA inflammatory infiltrates and in aneurysmal lesions of AngII-infused ApoE−/− mice.
- Bempedoic acid, an ACLY inhibitor, protected mice against AAA formation, reducing inflammatory-destructive vascular remodeling.
- Findings link immunometabolic ACLY signaling in myeloid/lymphoid cells to aneurysm biology, supporting drug repurposing.
Methodological Strengths
- Human tissue corroboration alongside an established murine aneurysm model
- Target engagement of a clinically used metabolic enzyme (ACLY)
Limitations
- Preclinical mouse data; clinical dosing, safety, and efficacy for AAA remain unknown
- Potential confounding lipid effects of bempedoic acid not fully disentangled from anti-inflammatory actions
Future Directions: Dose-finding, biomarker-guided early-phase clinical trials in small AAA; mechanistic studies dissecting myeloid vs lymphoid ACLY contributions and lipid-independent effects.
Abdominal aortic aneurysm (AAA) is a prevalent degenerative disease characterized by an exacerbated inflammation and destructive vascular remodeling. Unfortunately, effective pharmacological tools for the treatment of this disease remain a challenge. ATP-citrate lyase (ACLY), the primary enzyme responsible for acetyl-CoA biosynthesis, is a key regulator of inflammatory signaling in macrophages and lymphocytes. Here, we found increased levels of the active (phosphorylated) form of ACLY (p-ACLY) in the inflammatory infiltrate of AAA from patients and in aneurysmal lesions from angiotensin II (Ang II)-infused apolipoprotein E-deficient mice (ApoE...