Visualisation and Analysis of Speech Production with Electropalatography

Area of Science:

• Linguistics
• Speech Science
• Biomedical Engineering

Background:

• Speech production involves complex physiological processes.
• Electropalatography (EPG) visualizes tongue-palate contact during speech.
• Analyzing EPG data, specifically articulation asymmetry, is crucial for phonetic and phonological studies.

Purpose of the Study:

• To present a novel open-source toolbox for analyzing electropalatography data.
• To enable detailed investigation of left-right asymmetry in tongue-palate contact.
• To provide tools for analyzing 3D computed tomography data of artificial palates.

Main Methods:

• Development of Matlab routines for EPG data visualization and analysis.
• Implementation of algorithms to quantify spatial and temporal patterns of tongue-palate contact.
• Creation of a pipeline for segmenting and analyzing 3D CT scans of artificial palates.

Main Results:

• The developed toolbox facilitates the analysis of palatograms and associated acoustic signals.
• Quantitative measures of left-right asymmetry in tongue-palate contact were obtained.
• The CT analysis pipeline provided insights into anatomical and manufacturing contributions to asymmetry.

Conclusions:

• The open-source toolbox enhances the study of speech production and articulation.
• The tools offer valuable insights into factors influencing speech asymmetry, such as anatomy and handedness.
• This work supports future research in electropalatography and speech science.