A Survey on Cardiovascular Care Using a Hybrid Ml/Dl Framework for Early Detection and Real-Time Monitoring

Abstract

This survey has provided a comprehensive review of conventional ML methods, deep learning architectures, and the hybrid frameworks that are applied to cardiovascular care. The conventional classifiers, such as support vector machines, random forests, and the gradient boosting methods, have shown efficiency and interpretability, which are specifically for structured data. Deep learning models, which include the convolutional neural networks, recurrent neural networks, graph neural networks, and generative models, perform well at capturing the complex spatial, temporal, and multi-modal patterns, which enhances the disease detection and prediction accuracy. Hybrid ML/DL frameworks have combined the feature extraction and classification across multiple data types, which achieves superior robustness and clinical applicability. The real-time monitoring and clinical decision support systems have exploited the edge computing, low-latency inference pipelines, and the wearable or telemetry devices, which allow a timely intervention and continuous patient surveillance. The survey also addresses the challenges such as data heterogeneity, limited multi-modal datasets, computational complexity, interpretability, and regulatory compliance. Emerging trends have focused on multi-modal fusion strategies, explainable AI, synthetic data augmentation, personalized longitudinal modeling, and federated learning as future directions to improve the generalization, reliability, and real-world deployment. By the consolidation of methodologies, datasets, evaluation metrics, and clinical case studies, this survey has provided a roadmap for researchers and clinicians that seeks to develop an explainable, efficient, and high-performance ML/DL framework for cardiovascular healthcare.