Detection of Cardiovascular Disease Risk Using Heart Rate Variability and Machine Learning: A Comprehensive Analysis

Abstract

Cardiovascular diseases (CVDs) remain the leading cause of global mortality, accounting for approximately 17.9 million deaths annually. Heart rate variability (HRV), a noninvasive marker reflecting autonomic nervous system function, has emerged as a promising predictor of cardiovascular events. Recent advances in artificial intelligence and machine learning have revolutionized HRV analysis, enabling more accurate detection and prediction of cardiovascular diseases. This comprehensive literature review synthesizes 19 seminal studies published between 2017 and 2025, examining the methodologies, machine learning algorithms, datasets, HRV features, and performance outcomes of ML-based cardiovascular disease detection systems. The review is organized chronologically to trace the evolution of techniques from conventional statistical methods to sophisticated deep learning architectures. Key findings reveal that ensemble methods, deep learning algorithms (particularly CNNs and LSTMs), and hybrid models consistently outperform traditional machine learning approaches, achieving accuracies exceeding 95% in many studies. The integration of multiple HRV features—including time-domain, frequency-domain, and non-linear measures—with clinical parameters has demonstrated superior predictive performance. However, challenges remain in standardizing HRV extraction protocols, addressing dataset imbalances, and validating models in real-world clinical settings. This review provides essential insights for researchers and clinicians seeking to leverage HRV-based machine learning systems for cardiovascular risk stratification and early disease detection.