Published Paper Details:

The increasing reliance on Medical Internet of Things (MIoT) devices for patient monitoring and healthcare delivery has introduced significant security challenges, necessitating advanced threat detection mechanisms. This review provides a comprehensive analysis of deep learning approaches for MIoT threat detection, while also considering complementary techniques such as machine learning, anomaly detection, blockchain-based security, hybrid models, and cloud-edge integration. Traditional machine learning methods like SVM, Decision Trees, Random Forests, and k-NN are effective for structured data but struggle with complex healthcare datasets, whereas deep learning models such as CNNs, RNNs, and LSTMs excel in feature extraction and sequential data analysis, achieving higher accuracy in detecting malicious activities. Anomaly detection techniques, including autoencoders and clustering methods, offer the ability to identify zero-day attacks but often suffer from false-positive rates. Blockchain integration enhances trust, transparency, and data integrity, though at the cost of latency and energy efficiency. Hybrid frameworks that combine ML/DL with anomaly detection or blockchain provide a balanced approach to security, and cloud-edge computing synergy further improves scalability and real-time responsiveness. Collectively, these approaches highlight the growing importance of deep learning in enabling secure, reliable, and intelligent threat detection in Medical IoT environments.