MEDICALHARM - A Threat Modeling designed for Modern Medical Devices

Modern Medical Devices (MMDs) are a rapidly growing field of medical technology, and recent advances have allowed them to monitor and manage patients’ health remotely. As these devices become more connected in order to enhance the delivery of patient care, the concerns surrounding security, privacy, and safety are also increasing. To effectively address these concerns, "shift-left security"-which involves addressing security risks as early as possible-is becoming increasingly important. To facilitate it, threat modeling must be implemented as the first step. While various threat modeling methodologies exist, MMDs need a tailored one that can take into account the safety of patients and the complexity of a typical Medical Device (MD), which contains multiple sensors and actuators. Therefore, we present a new threat modeling methodology -MEDICALHARM- tailored to identifying threats in MMD systems. MEDICALHARM delivers a holistic approach by combining threat and risk analysis under the same scheme. It specifically articulates safety threats along with security and privacy threats. Furthermore, it offers an algorithmic scheme to enable non-security experts (engineers and developers) to easily participate in the threat modeling process. To illustrate its benefits, we perform a threat modeling exercise using MEDICALHARM on a Deep Brain Stimulation device and provide an exhaustive threats document. We then compare the results of this exercise with another threat model scheme (STRIDE) to demonstrate MEDICALHARM’s efficiency in threat identification.