Analysis of auditory evoked potentials by magnitude-squared coherence

Area of Science:

• Neuroscience
• Auditory Neuroscience
• Signal Processing

Background:

• Evoked potentials are typically analyzed in the time domain.
• Frequency-domain analysis, like power spectral density, cannot differentiate signal from noise.
• Coherence function offers a method to quantify signal-to-noise ratio in frequency-domain analysis.

Purpose of the Study:

• To present and evaluate the coherence analysis method for human scalp responses to auditory stimuli.
• To demonstrate the utility of coherence analysis in understanding the spectral characteristics of auditory evoked potentials.
• To assess the threshold sensitivity improvements offered by coherence analysis compared to time-domain methods.

Main Methods:

• Coherence analysis was applied to human scalp responses.
• Stimuli included clicks and amplitude-modulated tones.
• Statistical significance was determined by comparing coherence values to critical thresholds.

Main Results:

• Coherence analysis provided insights into the spectral character of auditory evoked potentials.
• This method aids in defining optimal filter characteristics for signal processing.
• Statistically significant responses were detected at lower stimulus intensities using coherence analysis than with time-domain waveform inspection.

Conclusions:

• Coherence analysis is a valuable tool for analyzing auditory evoked potentials.
• It improves the detection sensitivity and provides spectral information about the response.
• This method enhances the understanding and analysis of the human auditory system's response to sound.