Vulnerabilities in Machine Learning-Based Voice Disorder Detection Systems

TL;DR

This paper investigates the security vulnerabilities of machine learning systems used for voice disorder detection, demonstrating how various attack methods can compromise their reliability and highlighting the need for improved security measures.

Contribution

It introduces an analysis of attack strategies against voice disorder detection models, revealing their susceptibility and emphasizing the importance of securing healthcare AI systems.

Findings

Adversarial and evasion attacks can significantly reduce detection accuracy

Pitching attacks can alter classification outcomes

Certain attack methods are more effective against current models

Abstract

The impact of voice disorders is becoming more widely acknowledged as a public health issue. Several machine learning-based classifiers with the potential to identify disorders have been used in recent studies to differentiate between normal and pathological voices and sounds. In this paper, we focus on analyzing the vulnerabilities of these systems by exploring the possibility of attacks that can reverse classification and compromise their reliability. Given the critical nature of personal health information, understanding which types of attacks are effective is a necessary first step toward improving the security of such systems. Starting from the original audios, we implement various attack methods, including adversarial, evasion, and pitching techniques, and evaluate how state-of-the-art disorder detection models respond to them. Our findings identify the most effective attack strategies, underscoring the need to address these vulnerabilities in machine-learning systems used in the healthcare domain.