Trustworthy and Ethical AI for Intrusion Detection in Healthcare IoT (IoMT) Systems: An Agentic Decision Loop Framework

🧠 Project Title<b>Trustworthy and Ethical AI for Intrusion Detection in Healthcare IoT (IoMT) Systems: An Agentic Decision Loop Framework</b>📋 OverviewThis repository contains the official code, datasets, and configuration setup for the paper submitted to <i>Springer’s Journal of Healthcare Informatics Research (JHIR)</i>.<br>The study presents a multi-agent intrusion detection architecture that integrates:A supervised flow-based detectorA Deep Q-Network (DQN) triage agentA NIST AI RMF–aligned ethical rule engineThe framework enables <b>trustworthy, safe, and context-aware intrusion detection</b> in healthcare IoT environments (IoMT).🏗️ Repository Structure<pre><pre>agentic-ethical-ids-healthcare/<br>│<br>├── src/ # Source code for model, rule engine, and agent<br>│ ├── train_agent.py<br>│ ├── ethical_engine.py<br>│ ├── detector_model.py<br>│ └── utils/<br>│<br>├── data/ # Links or sample data subsets<br>│ ├── CIC-IoMT-2024/ <br>│ └── CSE-CIC-IDS2018/<br>│<br>├── notebooks/ # Jupyter notebooks for training and analysis<br>│<br>├── models/ # Pretrained model checkpoints (.pth, .pkl)<br>│<br>├── results/ # Evaluation outputs and figures<br>│<br>├── requirements.txt # Python dependencies<br>├── LICENSE # MIT License for open research use<br>└── README.md # Project documentation<br></pre></pre>⚙️ Setup and InstallationClone the repository and set up your environment:<pre><pre>git clone https://github.com/ibrahimadabara01/agentic-ethical-ids-healthcare.git<br>cd agentic-ethical-ids-healthcare<br>python -m venv venv<br>source venv/bin/activate # On Windows: venv\Scripts\activate<br>pip install -r requirements.txt<br></pre></pre>📊 DatasetsThis project uses three datasets:DatasetPurposeSource<b>CIC-IoMT 2024</b>Primary IoMT intrusion detection datasetCanadian Institute for Cybersecurity<b>CSE-CIC-IDS2018</b>Domain-shift evaluationCIC Dataset Portal<b>MIMIC-IV (Demo)</b>Clinical context signalsPhysioNet⚠️ Note: All datasets are publicly available. The MIMIC-IV Demo contains only de-identified data.🚀 How to Reproduce ResultsRun the full pipeline (training + evaluation):<pre><pre>python src/train_agent.py --config configs/agentic_ids.yaml<br></pre></pre>This script:Trains the supervised flow-based detector on CIC-IoMT 2024Fine-tunes the DQN triage agentEvaluates under domain-shift using CSE-CIC-IDS2018Computes Ethical Compliance Rate (ECR), False Escalation Rate (FER), and CAS metrics📈 Key MetricsMetricDescription<b>Accuracy</b>Correct classification rate across all flows<b>F1-Score (Weighted)</b>Balanced measure of precision and recall<b>Ethical Compliance Rate (ECR)</b>Percentage of actions consistent with governance rules<b>False Escalation Rate (FER)</b>Proportion of overreactions (false alarms)<b>Contextual Adaptation Score (CAS)</b>Robustness under domain-shift📘 CitationIf you use this repository, please cite:<pre><pre>Adabara, I. M., et al. (2025). Trustworthy and Ethical AI for Intrusion Detection in Healthcare IoT (IoMT) Systems: An Agentic Decision Loop Framework. Journal of Healthcare Informatics Research, Springer.<br></pre></pre>🔒 Ethical ComplianceAll experiments comply with PhysioNet and HIPAA de-identification standards.<br>The MIMIC-IV Demo dataset was used under credentialed access and contains no PHI.🧾 LicenseThis project is released under the <b>MIT License</b>, allowing free use for research and educational purposes.