Application of Deep Learning in Medical Cyber-Physical Systems

The integration of IoT devices to healthcare sector has enabled remote monitoring of patient data and delivery of suitable diagnostics whenever required. Because of the rapid advancement in embedded software and network connectivity, Cyber physical systems (CPS) have been widely used in the medical industry to provide top-notch patient care in a variety of clinical scenarios because of the quick advancements in embedded software and network connectivity. Due to the heterogeneity of the medical devices used in these systems, there is a requirement for providing efficient security solutions for these intricate environments. Any alteration to the data could have an effect on the patient’s care, which may lead to accidental deaths in an emergency. Deep learning has the potential to offer an efficient solution for intrusion detection because of the high dimensionality and conspicuous dynamicity of the data involved in such systems. Therefore, in this study, a deep learning-assisted Attack Detection Framework has been suggested for safely transferring healthcare data in medical cyber physical systems. Additionally, the efficacy of the suggested framework in comparison to various cutting-edge machine and ensemble learning techniques has been assessed on healthcare dataset consisting of sixteen thousand records of normal and attack data and the experimental findings indicate that the suggested framework offers promising outcomes when compared with the state-of-the-art machine learning and ensemble learning approaches.