Application of Adversarial Examples to Physical ECG Signals

TL;DR

This paper investigates the vulnerability of ECG-based cardiac diagnosis systems to adversarial attacks by creating and physically injecting adversarial beats, demonstrating their potential to manipulate diagnosis results in real-world scenarios.

Contribution

It introduces a novel method to generate adversarial ECG beats and evaluates their effectiveness in physical environments, highlighting security risks in medical diagnostics.

Findings

Adversarial beats successfully manipulated diagnosis 3-5 times out of 40 attempts

Physical injection of adversarial signals is feasible and effective

First study to evaluate ECG adversarial attacks in real-world settings

Abstract

This work aims to assess the reality and feasibility of the adversarial attack against cardiac diagnosis system powered by machine learning algorithms. To this end, we introduce adversarial beats, which are adversarial perturbations tailored specifically against electrocardiograms (ECGs) beat-by-beat classification system. We first formulate an algorithm to generate adversarial examples for the ECG classification neural network model, and study its attack success rate. Next, to evaluate its feasibility in a physical environment, we mount a hardware attack by designing a malicious signal generator which injects adversarial beats into ECG sensor readings. To the best of our knowledge, our work is the first in evaluating the proficiency of adversarial examples for ECGs in a physical setup. Our real-world experiments demonstrate that adversarial beats successfully manipulated the diagnosis results 3-5 times out of 40 attempts throughout the course of 2 minutes. Finally, we discuss the overall feasibility and impact of the attack, by clearly defining motives and constraints of expected attackers along with our experimental results.