ERP Signals During Speech Articulation: Does Auditory Feedback Mask Other Ongoing Cognitive-motor Processes?

Area of Science:

• Neuroscience
• Cognitive Science
• Speech Science

Background:

• Analyzing Event-Related Potentials (ERPs) during speech production is crucial for understanding speech planning.
• ERPs from speech production are often contaminated by brain signals from auditory feedback.
• Current signal separation algorithms struggle to differentiate these time-locked signals.

Purpose of the Study:

• To investigate the efficacy of the multi-channel Wiener filter (MWF) in separating auditory feedback ERPs from speech production ERPs.
• To compare the effects of MWF filtering, signal subtraction, and original production ERPs on microstate analysis.
• To determine if auditory feedback removal is necessary for accurate speech production ERP analysis.

Main Methods:

• Participants produced pseudowords, then listened to recordings of their own speech.
• Multi-channel Wiener filter (MWF) was used to estimate and remove auditory ERPs from production ERPs.
• Microstate analysis was performed on original, MWF-filtered, and subtraction-based ERPs, comparing conditions.

Main Results:

• The subtraction method altered topography consistency and potentially removed relevant ERP signals.
• MWF-filtered ERPs produced results comparable to the original production ERPs.
• Auditory feedback filtering might be unnecessary when signals are time-locked to speech production onset.

Conclusions:

• The multi-channel Wiener filter (MWF) is a viable method for processing speech production ERPs.
• Signal subtraction may distort the original ERP data.
• Auditory feedback does not necessarily need to be removed from production ERPs for accurate analysis when time-locked to vocal onset.