Automated Detection of Cybersecurity Intrusions in Healthcare Systems Using Several Approaches

Abstract

To ensure that patients are receiving the proper care, the healthcare data must be improved, real-time monitored, and accurate in illness detection. Thus, machine learning techniques are widely employed in Smart Healthcare Systems (SHS) to extract valuable features for tracking patient behaviors and forecasting various diseases from diverse and high-dimensional healthcare data. The kidneys gradually lose their functionality as a result of chronic kidney disease (CKD). It talks about a medical condition that damages the kidneys and has an impact on a person's overall health. In this study, recursive feature elimination (RFE) and multilayer perceptron’s are used to develop a model for identifying anomalies and cyber-attacks (MLP). Experimental data are used to evaluate the suggested MLP model's performance. Recall, precision, accuracy, and F1-score are only a few of the performance metrics used to forecast patient activities. When compared to the RFE technique, the recommended strategy provides the highest levels of accuracy, precision, recall, and F1-score. Specifically, 98.56% recall, 98.13% F1-score, 98.76 accuracy, and 98.93% precision are obtained using the proposed MLP technique. When the outcomes were compared to recent state-of-the-art and machine learning algorithms from recent times, they performed better.