SpeechMark Acoustic Landmark Tool: Application to Voice Pathology.

Area of Science:

• Speech Science
• Acoustic Phonetics
• Voice Disorders

Background:

• The relationship between voice pathology and speech intelligibility remains understudied.
• Acoustic analysis of speech signals is crucial for understanding vocal production.
• Existing methods may not efficiently capture subtle acoustic variations linked to pathology.

Purpose of the Study:

• To explore the application of landmark analysis in evaluating intelligibility-based measures of vocal pathology.
• To develop a toolset for fast, automatic acoustic analysis of speech.
• To bridge the gap between voice pathology research and acoustic signal analysis.

Main Methods:

• Utilizing landmark-based software for automatic analysis of speech acoustics.
• Applying Stevens' paradigm of landmarks to extract articulatory event information.
• Generating summary statistics for speech acoustic measures.

Main Results:

• Landmark analysis provides a suitable method for evaluating intelligibility-based measures.
• The developed toolset enables fast and automatic acoustic assessment.
• Demonstrated potential for identifying subtle acoustic differences related to vocal pathology.

Conclusions:

• Landmark analysis is a promising approach for the objective evaluation of vocal pathology.
• Automatic acoustic analysis tools can enhance the study of voice disorders and intelligibility.
• Further research can refine these methods for clinical applications.