Resiliency analysis and modeling for real-time cardiovascular diagnostic devices

摘要

To address the increasing severe challenges with regard to the reliability of medical devices, in this study we propose to investigate the resilience and vulnerability of cardiovascular diagnostic devices as a case study, by using a formal fault injection approach. We develop a generalized model of the ECG-based heart disease detection system using MATLAB/Simulink, which consists of heartbeat detection and classification. Furthermore, we propose a configurable fault generator (FG) to emulate a wide variety of accidential faults occuring within the developed system, to help us explore and understand the most critical data-paths and processing modules, where any faults or unexpected perturbations will cause dramatic impacts on the system accuracy and performance, or even result in system crash in real case.