A Hybrid Deep Learning Approach for Accurate Network Intrusion Detection Using Traffic Flow Analysis in IoMT Domain

The advancement of technology has led to the rise of the Internet of Medical Things (IoMT), revolutionizing the healthcare sector by integrating medical systems and devices. This integration, however, introduces various cybersecurity risks. This paper presents a machine learning (ML) approach to enhance IoMT security. We explore the application of ML in identifying and predicting cyber threats in the IoMT landscape. Traditional cybersecurity measures often fall short in the face of evolving threats. Thus, we introduce a hybrid deep learning model, combining Random Forest's feature selection with deep neural network's pattern recognition for intrusion detection. This model achieves an accuracy of 94.48%, precision of 84.62%, recall of 94.28%, and F1-score of 92.89%. Using the ECU_IoHT dataset for validation, our ML-centric framework outshines conventional methods in threat detection and minimizes false positives. This research underscores the potential of ML in fortifying IoMT, ensuring patient privacy, system integrity, and continuous healthcare delivery.