Faculty Advisor

Krishna Kumar Venkatasubramanian

Faculty Advisor

Craig C. Shue

Abstract

"Security for medical devices has gained some attractions in the recent years following some well- publicized attacks on individual devices, such as pacemakers and insulin pumps. This has resulted in solutions being proposed for securing these devices, usually in stand-alone mode. Medical devices are however becoming increasingly interconnected and interoperable as a way to improve patient safety, decrease false alarms, and reduce clinician cognitive workload. Given the nature of interoperable medical devices (IMDs), attacks on IMDs can have devastating consequences. This work outlines our effort in understanding the threats faced by IMDs, an important first step in eventually designing secure interoperability architectures. A useful way of performing threat analysis of any system is to use attack trees. Attack trees are conceptual, multi-leveled diagrams showing how an asset, or target, might be attacked. They provide a formal, methodical way of describing the threats to a system. Developing attack trees for any system is however non-trivial and requires considerable expertise in identifying the various attack vectors. IMDs are typically deployed in hospitals by clinicians and clinical engineers who may not posses such expertise. We therefore develop a methodology that will enable the automated generation of attack trees for IMDs based on a description of the IMD operational workflow and list of safety hazards that need to be avoided during its operation. Additionally, we use the generated attack trees to quantify the security condition of the IMD instance being analyzed. Both these pieces of information can be provided by the users of IMDs in a care facility. The contributions of this paper are: (1) a methodology for automated generation of attack trees for IMDs using process modeling and hazard analysis, and (2) a demonstration of the viability of the methodology for a specific IMD setup called Patient Controlled Analgesia (PCA- IMD), which is used for delivering pain medication to patients in hospitals."

Publisher

Worcester Polytechnic Institute

Degree Name

MS

Department

Computer Science

Project Type

Thesis

Date Accepted

2015-12-10

Accessibility

Unrestricted

Subjects

Attack Trees, Interoperable Medical Device System, Vulnerability Evaluation

Share

COinS