Blockchain-Integrated Bidirectional Long Short-Term Memory Network for Real-Time Intrusion Detection in Healthcare...

Area of Science:

• Cybersecurity
• Network Security
• Healthcare Technology

Background:

• Healthcare Internet of Medical Things (IoMT) systems face significant cybersecurity challenges, requiring robust intrusion detection systems (IDS) with forensic capabilities.
• Traditional IDSs often lack tamper-proof logging and effective post-incident investigation support, hindering accountability in healthcare environments.

Purpose of the Study:

• To develop a forensic-aware intrusion detection framework for healthcare IoMT systems.
• To integrate an Extended Bidirectional Long Short-Term Memory (BiLSTM) network with a permissioned blockchain for real-time detection, secure logging, and automated mitigation.

Main Methods:

• The framework employs data preprocessing, AQU-IMF-RFE feature selection, and an Extended BiLSTM model for temporal sequence modeling and attention-based learning.
• Intrusion events are recorded on a Proof-of-Authority blockchain via smart contracts for immutable logging and automated responses.
• The Extended BiLSTM model was rigorously evaluated on UNSW-NB15, CICIDS2017, and Bot-IoT datasets.

Main Results:

• The proposed system achieved high intrusion detection accuracy with low false-positive rates across benchmark datasets.
• The blockchain layer ensured tamper-resistant audit trails and enabled automated mitigation actions.
• The framework demonstrated effective forensic traceability and real-time response capabilities without significant computational overhead.

Conclusions:

• Integrating deep learning-based intrusion detection with blockchain-enabled forensic logging enhances the trustworthiness and accountability of healthcare IoMT cybersecurity.
• The developed framework offers a practical and deployable solution for securing sensitive healthcare data and connected medical devices.