Transforming the cybersecurity space of healthcare IoT devices using Deep Learning

The proliferating growth of IoT devices has created a great revolution in the lives of mankind especially in the healthcare sector. On the other side, there is an inclining trend of cyberattacks on healthcare IoT devices as the clinical data is very sensitive and not many healthcare organizations impart stringent security protocols on Internet of Medical Things (MIoT) devices. The distinct property of IoT devices and their resource constrained nature limits the effectiveness of existing security solutions. To combat this issue, the present research work proposes a Deep Gated Recurrent Unit (D-GRU) as an AI based framework to predict the onset of attacks in MIoT devices in Critical Care Units (CCU). The proposed model explores the various network and communication parameters of the patient and environment monitoring devices employed in CCU to detect the attack. This framework is lightweighted and hence can be easily deployed in healthcare sectors, which would eventually transform the cybersecurity landscape of MIoT devices. The proposed model is used to predict the onset of cyberattacks in open source, synthetic IoT healthcare data generated through IoT Flock tool. The intricate performance analysis of the model on various classification metrics shows that the model is effect in detecting the attacks in MIoT.