Advanced Deep Learning Frameworks for Cyber Security in IoT-Based Healthcare
Usharani Bhimavarapu · 2025
The rapid adoption of IoT-based healthcare devices has revolutionized the medical industry by enabling real-time monitoring and data collection, yet it raises significant privacy and security concerns. These devices are vulnerable to various cyber-attacks, including man-in-the-middle attacks, spoofing, and data injection, which can compromise sensitive patient information and disrupt critical healthcare operations. To address these challenges, this study explores the application of deep learning techniques, particularly ResNet-50, to classify unforeseen cyber-attacks in IoT-based healthcare systems. The WUSTL-EHMS-2020 dataset, containing biometric data and network flow metrics, is utilized for experimentation. Preprocessing steps like normalization and feature engineering ensure compatibility with the model. ResNet-50, with its residual learning mechanism, effectively extracts complex patterns and generalizes to previously unseen attacks. The model's performance is evaluated using accuracy, precision, recall, and F1-score, demonstrating its ability to identify cyber-attacks while minimizing false positives and negatives. This research underscores the importance of advanced deep learning frameworks in enhancing the security and reliability of IOT-based healthcare systems.