Scholarlink

<p>The Internet of Medical Things (IoMT) plays a pivotal role in the healthcare industry, fostering remote patient monitoring and generating alerts in emergencies, thereby improving patient health and driving innovation in today’s time. However, these IoMT devices and sensors are vulnerable to cybersecurity threats such as device tampering, information breaches, and patient security concerns. Since the healthcare industry is evolving and patient privacy is crucial, detecting and mitigating these cybersecurity attacks is critically essential. This manuscript presents an Explainable Artificial Intelligence (XAI)-driven hybrid Anomaly Detection System (ADS) in blockchain-oriented cyber-physical systems for the healthcare IoMT. An improved AI-based model with an encoded and normalized feature set is proposed. The developed model is applied to the CICIoMT 2024 dataset, which has 45 features and detects 19 attacks, which majorly include Distributed Denial of Service (DDoS), Denial of Service (DoS), Reconnaissance (Recon), Message Queuing Telemetry Transport (MQTT), and spoofing from the system. Blockchain-based data authentication and user registration have been integrated for decentralized security using Nik 512’s cryptographic strength and off-chain storage for optimized data management. The proposed system has been evaluated on several parameters. The system demonstrates a detection rate with an accuracy rate of 99. 9%, a precision rate of 99. 78%, a recall rate of 99.89%, an F1 score of 98%, and a specificity of 99. 98% for 2, 6, and 19 class classification, showcasing the efficiency of the proposed optuna-based XGBoost model compared to the existing methods. In addition, XAI methods such as SHAP and LIME have been used to interpret the predictions generated by the proposed model.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing IoMT security: a novel blockchain-based anomaly detection framework leveraging explainable AI

  • 期刊论文

  • 发表日期：

    2026年1月6日

查看全文
  • Aashima Sharma,
  • Swastik Sharma,
  • Simarjit Kaur

摘要

The Internet of Medical Things (IoMT) plays a pivotal role in the healthcare industry, fostering remote patient monitoring and generating alerts in emergencies, thereby improving patient health and driving innovation in today’s time. However, these IoMT devices and sensors are vulnerable to cybersecurity threats such as device tampering, information breaches, and patient security concerns. Since the healthcare industry is evolving and patient privacy is crucial, detecting and mitigating these cybersecurity attacks is critically essential. This manuscript presents an Explainable Artificial Intelligence (XAI)-driven hybrid Anomaly Detection System (ADS) in blockchain-oriented cyber-physical systems for the healthcare IoMT. An improved AI-based model with an encoded and normalized feature set is proposed. The developed model is applied to the CICIoMT 2024 dataset, which has 45 features and detects 19 attacks, which majorly include Distributed Denial of Service (DDoS), Denial of Service (DoS), Reconnaissance (Recon), Message Queuing Telemetry Transport (MQTT), and spoofing from the system. Blockchain-based data authentication and user registration have been integrated for decentralized security using Nik 512’s cryptographic strength and off-chain storage for optimized data management. The proposed system has been evaluated on several parameters. The system demonstrates a detection rate with an accuracy rate of 99. 9%, a precision rate of 99. 78%, a recall rate of 99.89%, an F1 score of 98%, and a specificity of 99. 98% for 2, 6, and 19 class classification, showcasing the efficiency of the proposed optuna-based XGBoost model compared to the existing methods. In addition, XAI methods such as SHAP and LIME have been used to interpret the predictions generated by the proposed model.