Treating a different kind of patient: curing security weaknesses in digital health systems of the future

Abstract

Technological innovations are changing the world we live in, but in no sector they can have a greater impact on our lives than in the area of digital healthcare. The potential is immense: from AI-driven drug discovery to pervasive sensing to predict and prevent diseases and health conditions, the promise of technology-enabled medicine is that of better diagnoses, more effective treatments, improved prognoses, and ultimately longer and healthier lives.While technological advances in healthcare are being introduced at an ever increasing pace, they face an existential risk for their eventual adoption and public acceptance: they do not currently embed advanced cyber security mechanism that will be capable of confronting the complex cyber threats landscape of the future. In fact, while advanced cryptographic and security techniques such as secure multi-party computation, federated learning and root of trust are edging closer to deployment in the real world, they have largely not been adopted in the medical and healthcare domain. In this work, we argue that such adoption, enabled by stronger collaboration and deeper mutual knowledge between the healthcare and cyber security research community, is crucial for the success and safety of medical devices and sensors. In particular, we discuss the case of sensors, and their security, as well as the privacy of the data they collect.