An ML Approach to Analyze Cyberthreats and Vulnerabilities in the Healthcare Ecosystem

Hospitals are beginning to face a threat from cyberattacks, jeopardizing patients’ lives. Medical cyberphysical systems are a major source of potential vulnerabilities (MCPS). The entire computing paradigm has changed due to advances in information and communication technology (ICT). These advancements lead to many new communication channels, including the Internet of Things (IoT). With the adoption of major enabler technologies like blockchain, artificial intelligence (AI), Internet of Things (IoT), and next-generation wireless networks (5G/6G), the healthcare industry has undergone a massive upheaval. This will support remote patient examination and monitoring, containment of the epidemic, exposure reduction, area disinfection, provision of food and medication, awareness-building, and vital sign monitoring for early detection. The patient-provider interaction is strengthened by the use of IoT technologies in the healthcare industry. It makes it possible for portable devices, known as sensor nodes, to quickly gather biometric data from patients’ bodies such as temperature, blood pressure, oxygen saturation, and so on and send it to other sensor nodes or straight to healthcare providers, including doctors, pharmacists, and labs. The sensors include memory, processor, and attribute and data profiles. The attribute profile specifies the manufacturer, kind, measurement range, manufacturing date, and sensor position. The data profile handles the data format that the sensors are producing. The proposed logistic regression classifier's accuracy, F1 measure, recall, and precision are achieved at 96.29, 94.38, 93.67, and 95.42%. Greater performance is obtained as the training set size grows. The system works better with data augmentation added than without it. The existing systems used for comparison are SVM, CNN, and ANN algorithms. To assess the effectiveness of the suggested approach, Matlab 2013A is utilized to create a machine learning algorithm in healthcare ecosystem.