Artificial intelligence for risk prediction in atherosclerotic cardiovascular disease: A narrative review of advances, validation challenges, and clinical translation (2020-2026).
Aug 2026· Progress in cardiovascular diseases· 0 citations· 35 references
Medicine
TL;DR
Future work should prioritize calibration, robustness, transportability, fairness, interpretability, regulatory clarity, workflow integration, and prospective evidence of decision impact or post-deployment benefit, as well as major barriers remain.
Abstract
Since 2020, artificial intelligence (AI) has been increasingly applied to atherosclerotic cardiovascular disease (ASCVD) risk prediction. This structured narrative review with systematic evidence mapping summarizes literature (2020-2026) examining study design, data sources, model architectures, multimodal fusion, model development and validation, performance evaluation, subgroup applications, and implementation barriers. Overall, 126 studies informed the review; 93 provided sufficient information for structured extraction, including prevention setting, exact input variables, comparator scores, validation strategies, discrimination, calibration, dominant model architecture, endpoint category, foundation-model or pretrained-model status, regulatory status, and implementation features. AI-based models may offer modest but clinically meaningful gains over conventional risk equations, especially with multimodal or longitudinal data. Among the 93 studies, traditional machine learning accounted for 83 (89%), deep learning for 7 (8%), and multimodal fusion for 3 (3%). Endpoint definitions were heterogeneous (23% ASCVD-specific; 63% expanded MACE composites). Among these studies, no large language model or federated learning was used for risk prediction. Appropriate comparators should now include contemporary equations such as PREVENT, rather than only legacy tools. Major barriers remain, including limited external validation, performance attenuation, data and algorithmic bias, limited interpretability, inconsistent reporting of calibration, fairness, and clinical utility, unclear regulatory status, and limited prospective evidence. Future work should prioritize calibration, robustness, transportability, fairness, interpretability, regulatory clarity, workflow integration, and prospective evidence of decision impact or post-deployment benefit. The central question is not only whether AI can detect complex patterns, but whether such models can be trusted, implemented, and shown to advance preventive cardiology in real-world settings.
Introduction: Cardiovascular risk prediction remains challenging, particularly in patients with intermediate risk, mixed dyslipidemia, elevated lipoprotein(a), or variable lipid profiles. Conventional risk calculators may not fully capture nonlinear relationships among lipid, clinical, imaging, and longitudinal data. Objectives: This narrative review summarizes evidence on artificial intelligence (AI)-based cardiovascular risk assessment, focusing on lipid profile-based and multimodal models incorporating lipid-related variables. Methods: PubMed/MEDLINE, Scopus, and Google Scholar were searched for English-language articles published up to January 2026. Original studies, reviews, and relevant clinical guidelines addressing AI-based cardiovascular risk models, lipid-related predictors, and clinically applicable approaches were considered. Results: Lipid profile-based AI models may identify lipid phenotypes and lipid-related patterns associated with increased cardiovascular risk, while multimodal models have shown improved performance in selected datasets. However, the reviewed studies address heterogeneous tasks, including phenotype classification, cardiovascular event prediction, mortality prediction, patient trajectory modeling, and absolute risk estimation. Most evidence remains retrospective, with limited external validation, calibration assessment, and clinical utility data. Conclusions: AI-based models may support cardiovascular risk assessment, but routine implementation requires prospective validation, standardized evaluation, calibration, explainability, and clinical impact studies.
P. Piłat, Radosław Dutczak, Mariusz Gąsior et al.· Journal of Clinical Medicine· 0 citations
Evidence is provided that ensemble-based frameworks currently offer the most effective balance between predictive accuracy, robustness, and clinical feasibility, and future research should emphasize multi-center external validation and explainable AI frameworks.
Marium Shaikh, Hanmant Fadewar· International Journal For Mu...· 0 citations
It was demonstrated that machine learning (ML) and deep learning (DL) models consistently outperformed conventional cardiovascular risk prediction tools, achieving area under the receiver operating characteristic curve (AUC) values ranging from 0.80-0.99 across various cardiovascular conditions.
N. Muruganandan, P. Prathiba, Khyati Rajeshkumar Patel et al.· International Journal of Res...· 0 citations
Prediabetes is a highly prevalent intermediate metabolic state and a major public health target for preventing type 2 diabetes mellitus (T2DM). However, current approaches to prediabetes screening, risk stratification, and lifestyle management remain limited by inconsistent diagnostic definitions, incomplete case detection, heterogeneous progression risk, and the resource-intensive nature of conventional face-to-face prevention programmes. Artificial intelligence (AI), including machine learning, deep learning, explainable AI, and algorithm-driven digital interventions, is increasingly being explored as a tool to address these gaps. This narrative review used a structured search of PubMed, Web of Science Core Collection, and Embase for studies published from January 2010 to April 2026, selecting articles that evaluated AI-assisted or algorithm-driven approaches for prediabetes screening, progression risk prediction, or lifestyle intervention and reported relevant model performance, validation, or intervention outcomes. Current evidence indicates that AI-based models can improve discrimination beyond traditional risk scores by integrating routine clinical data, longitudinal electronic health records, continuous glucose monitoring profiles, wearable-derived behavioural signals, and emerging molecular biomarkers. Some externally validated models have shown clinically relevant performance for identifying individuals at high risk of progression and for guiding more targeted preventive strategies. In parallel, fully or semi-automated digital programmes delivered through mobile applications, web platforms, connected scales, and sensor-based feedback systems have demonstrated potential to support lifestyle change, improve engagement, and reduce reliance on labour-intensive counselling. Nevertheless, translation into routine care remains constrained by heterogeneity in prediabetes definitions, limited external validation across diverse populations, uncertain long-term effectiveness, geographical imbalance in evidence, privacy concerns, and the need for stronger governance frameworks. Overall, AI should be viewed as an assistive technology that may support earlier detection, more precise risk stratification, and scalable lifestyle management in prediabetes. Further multi-centre, prospective, and implementation-focused studies are needed to establish clinical utility, equity, safety, and cost-effectiveness.
Run-Xuan Chen, Min Song, Tong Xiang· Frontiers in Endocrinology· 0 citations
Background Stroke remains a leading cause of mortality, long-term disability, and healthcare expenditure worldwide, placing substantial strain on healthcare systems, particularly in low- and middle-income countries. Effective risk stratification can facilitate targeted prevention strategies, optimize resource allocation, and reduce avoidable hospitalizations. This study synthesizes existing evidence on the predictive performance of artificial intelligence (AI)-based models for stroke risk assessment through meta-analysis and explores their potential implications for healthcare system planning. Methods Studies were systematically retrieved from Web of Science (WoS), PubMed, and Scopus until 31 January 2025. The review followed the PRISMA 2020 guidelines. Area Under the Receiver Operating Characteristic Curve (AUC) values were extracted for each algorithm type and pooled using meta-analytic methods. Results Deep learning (DL) algorithms demonstrated favorable pooled discriminative performance (AUC: 0.955; 95% CI: 0.906–1.00, I2 = 85.75%), especially for imaging-based models. Sensitivity analysis modestly reduced heterogeneity (I2 from 85.75% to 61.77%). Substantial heterogeneity remained across study populations, healthcare settings, predictor characteristics, and validation strategies, limiting the generalizability of findings. Conclusions AI-based models, particularly DL approaches, demonstrate favorable predictive performance for stroke risk stratification. However, considerable methodological heterogeneity, limited external validation, and risk of bias reduce confidence in widespread clinical implementation. Future research should follow standardized reporting and validation frameworks, such as TRIPOD, to improve methodological rigor, transparency, and clinical applicability.
Raoof Nopour· Inquiry : a journal of medic...· 0 citations
This systematic review analyses the effectiveness of integrated AI‐based hypertension prediction and management systems those integrate real‐time oversight, interpretable risk prediction, customized lifestyle intervention, clinical decision support, and early warning procedure.
