Development and Evaluation of a Speech-to-Noise Ratio Feedback System

Area of Science:

• Neurology
• Speech-Language Pathology
• Biomedical Engineering

Background:

• Parkinson's disease (PD) frequently causes reduced vocal loudness, impairing speech intelligibility, particularly in noisy settings.
• Effective communication is crucial for individuals with PD, yet hypophonia presents a significant challenge.

Purpose of the Study:

• To develop and evaluate an assistive speech-to-noise ratio feedback (SNF) system for individuals with Parkinson's disease.
• To enhance vocal loudness and speech intelligibility in challenging acoustic environments.

Main Methods:

• Development of a real-time SNF system utilizing binaural microphones and a mobile application.
• Implementation of coherence-based own voice detection (OVD) and speech signal-to-noise ratio (SNR) estimation algorithms.
• Performance evaluation through electroacoustic and subjective testing across diverse environmental conditions.

Main Results:

• The OVD algorithm successfully distinguished the user's voice from background noise with appropriate threshold settings.
• Subjective tests confirmed that the SNF system significantly increased users' speech intensity during operation.
• The system demonstrated effectiveness in various environmental conditions.

Conclusions:

• The developed SNF system is a viable assistive tool for individuals with Parkinson's disease experiencing reduced vocal loudness.
• Real-time SNR feedback can effectively prompt users to increase speech intensity, thereby improving communication.
• This technology holds promise for enhancing the quality of life for people with PD.