IoMT Enabled Advanced Cyber-attack Detection Model (iACDM) for secure Healthcare System

The newest technology enabling patient life quality improvement functions through tele-health treatments that do not require specific time or space settings. Medical data exchanges over the Internet of Medical Things (IoMT) platform when using embedded products and sensors and actuators. Medical devices that fall under the category of Internet of Medical Things create safety and privacy concerns among patients because they are vulnerable to security threats. Security stands as an essential factor that determines the successful integration of IoMT technology into current healthcare systems. Security measures must be developed immediately to defend against peripheral network attacks of the IoMT system. Traditional approaches to protecting sensitive information have multiple weaknesses when used as data security methods. Attack detection for IoMT becomes more efficient based on the application of machine learning (ML) and artificial intelligence (AI). This paper demonstrates a new secure healthcare system design which implements iACDM as an IoMT-enabled advanced cyber-attack detection model. Cyber-attacks are detected by the proposed model which employs the Crow Search Optimization (CSO) algorithm to adjust its Cascaded Long Short Term Memory (CLSTM) model. The Application of CSO allows model weight and bias parameter adjustments which improves how CLSTM organizes incoming data categories. The current research utilizes isolation forest (IF) to remove outlier observations from the dataset. CSO application leads to substantial diagnostic result improvement in The CLSTM model. The proposed iACDM uses ToN-IoT dataset for conducting experiments and performance evaluations. The proposed model achieved superior attack detection accuracy than existing methods per research classification results.