A comprehensive survey on artificial intelligence and deep learning applications in cardiovascular disease: advances, challenges, and future directions
IET Conference Proceedings · 01.08.2026 · 2025 (63) · 599-608
DOI: 10.1049/icp.2026.2531 ↗Kurz gefasst
Fachartikel in der Zeitschrift IET Conference Proceedings, erschienen am 01.08.2026. Verfasst von Huma Tauseif und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Cardiovascular disease (CVD) remains the leading cause of global mortality, necessitating accurate, scalable, and timely diagnostic and prognostic solutions. In recent years, artificial intelligence (AI), particularly deep learning (DL), has shown remarkable promise in transforming cardiovascular care [1][2][3]. This survey presents a comprehensive review of the evolution and application of AI in cardiology, tracing developments from early neural networks to cardiologist-level diagnostic systems. Key milestones include the adoption of deep convolutional neural networks for electrocardiogram (ECG) interpretation, automated echocardiogram and cardiac MRI segmentation, and AI-driven decision support systems in interventional cardiology [4][5]. Studies reviewed demonstrate that AI models can match or exceed expert performance in arrhythmia detection, imaging analysis, and risk prediction, with accuracy metrics often surpassing traditional methods. The availability of large-scale clinical datasets such as MIMIC-III and CODE has played a pivotal role in enabling these advances [9]. Despite significant achievements, challenges remain in interpretability, data standardization, prospective validation, and clinical integration. This paper synthesizes current research, highlights performance benchmarks, and identifies open problems and future directions. The findings underscore AI's growing role in enhancing cardiovascular diagnostics and treatment, while emphasizing the need for responsible, human-centered deployment The convergence of artificial intelligence (AI) and deep learning (DL) has transformed cardiovascular medicine, allowing for computerized analysis of intricate physiological and imaging information [5][6]. AI methods now cover the entire spectrum of cardiovascular care — diagnosis, prognosis, and personalized therapy planning — ranging from ECG-based arrhythmia classification to multimodal cardiac imaging and retinal biomarker prediction. Recent advances like transformer-based ECG architectures, multimodal retinal and systemic biomarker models, and open cardiac MRI datasets like CMRxRecon have improved accuracy, generalization, and reproducibility [24]. Meta-analyses show that AI reaches expert-level performance for a multitude of tasks, whereas federated and explainable learning methods resolve data-sharing and transparency issues. This survey integrates recent progress, recognizes existing bottlenecks, and indicates translational avenues towards clinically deployable AI-powered cardiovascular systems.
Zitieren
APA
Tauseif, H., Vallem, S. R., & Khan, N. U. (2026). A comprehensive survey on artificial intelligence and deep learning applications in cardiovascular disease: advances, challenges, and future directions. IET Conference Proceedings, 2025(63), 599-608. https://doi.org/10.1049/icp.2026.2531
Vancouver
Tauseif H, Vallem SR, Khan NU. A comprehensive survey on artificial intelligence and deep learning applications in cardiovascular disease: advances, challenges, and future directions. IET Conference Proceedings. 2026;2025(63):599-608. doi:10.1049/icp.2026.2531
BibTeX
@article{Tauseif2026,
author = {Tauseif, Huma and Vallem, Shobha Rani and Khan, Nafis Uddin},
title = {A comprehensive survey on artificial intelligence and deep learning applications in cardiovascular disease: advances, challenges, and future directions},
journal = {IET Conference Proceedings},
year = {2026},
volume = {2025},
number = {63},
pages = {599--608},
doi = {10.1049/icp.2026.2531}
}
EUVIMED