Enhancing loMT Security with Deep Learning Based Approach for Medical IoT Threat Detection
Nadir Can Kavkas, Kazım Yıldız · 2025
The rapid spread of Internet of Medical Things (IoMT) systems makes the security of these systems critical. Although IoMT can improve patient care and optimize health care services, it also poses cyberattacks and data breaches. In this context, securing IoMT systems is important for patient privacy and system integrity. A deep learning-based approach is proposed in this study to detect cyberattacks that may occur in IoMT systems. Deep Neural Networks (DNN) and Long-Short Term Memory (LSTM) architectures are used on the CICIoMT2024 dataset. An Fl-score of 99% was achieved for binary classification and over 75% for categorical and multiclass classification tasks. In particular, the LSTM architecture exhibited high performance in detecting complex attack patterns. The experimental results provide new insights for future research and provide effective and scalable solutions to improve the security of IoMT systems.