As healthcare embraces the Internet of Medical Things (IoMT), the need for robust cybersecurity has never been more urgent. My project focuses on developing a highly efficient Intrusion Detection System (IDS) tailored to protect IoMT ecosystems from advanced cyber threats while addressing the constraints of resource-limited devices. This work was recognized for its impactful contributions to the field and accepted at the 2024 IEEE International Conference on Internet of Things and Intelligence Systems (IOTAIS).
Main Highlights of the Project:
Real-World Relevance: Leveraged the CICIoMT2024 dataset, which captures a wide range of malicious and benign IoMT network activity, to simulate real-world attack scenarios.
Cutting-Edge Machine Learning: Utilized State-of-the-Art models like Random Forest and XGBoost for their scalability and precision in handling large, imbalanced datasets.
Dimensionality Reduction: Implemented Recursive Feature Elimination with Cross-Validation (RFECV), reducing the dataset’s features by 44.45% without compromising performance.
Performance Excellence: Achieved a state-of-the-art weighted F1 score of 99.81%, significantly improving detection accuracy for both common and rare attack types.
Resource Efficiency: Designed the IDS to be lightweight and deployable on IoMT devices, ensuring seamless performance even with limited computational resources.
Rigorous Evaluation: Adopted stratified K-fold cross-validation to ensure the model performs reliably across diverse attack scenarios and imbalanced data.
Why It Matters:
This project bridges the gap between theoretical research and practical application, providing a scalable, efficient, and highly accurate cybersecurity solution for IoMT devices. By addressing evolving cyber threats and resource limitations, this IDS contributes to safeguarding patient data and enhancing the resilience of healthcare systems globally.