Lightweight machine learning for privacy-preserving and secure networked medical devices: The SEPTON project use cases

Abstract

Cybersecurity incidents are among the greatest concerns of businesses, government agencies, and private citizens today. In the modern world, the protection of data and information assets has become nearly as important as maintaining the security of physical assets. This creates the need for increased security implementations, leading to improved user acceptance of such applications and, as a consequence, to large-scale adoption of these technologies and full exploitation of their advantages. In healthcare, networked medical devices (NMDs), either referring to hospital medical equipment or wearables, can be vulnerable to security breaches, potentially affecting the safety and effectiveness of each device. In this work, we present the specific areas of recent machine learning research applied to networked medical device security, through the objectives of the Horizon Europe SEPTON research project. State-of-the-art lightweight machine learning approaches are highlighted and the corresponding challenges of cybersecurity applications, ranging from implantable devices to inter-institution medical data exchange use cases, are showcased.