Filtering and spectral characteristics of averaged auditory brain-stem response and background noise in infants.

Abstract

Filtering of electrode-recorded activity before averaging is used in evoked-potential measurements to reduce background noise under the assumption that unwanted spectral components will be suppressed without substantially altering neural activity. Desired filter characteristics depend on signal and noise spectra and filter choice can affect the validity and efficiency of ABR measurement. Spectral characteristics of the human auditory brain-stem response (ABR) change with subject age and state and with the level and spectrum of the stimulus. Brain-stem evoked potentials from infants comprise greater low-frequency spectral components than do adult ABRs. Physiologic noise can also change with age of the subject. This study compares the effects of 30- and 100-Hz high-pass filters on ABR amplitude and response Fsp in full-term infants. Stimuli used to elicit ABRs are moderate- and low-level clicks and 500-Hz tone bursts. Spectral characteristics of the averaged ABR and the no-stimulus background noise from full-term infants are also characterized. Results indicate that (1) energy in the infant ABR is concentrated below 100 Hz and (2) a high-pass recording filter of 30 Hz reveals a larger-amplitude ABR and enhances the overall signal-to-noise ratio as measured by Fsp as compared to a 100-Hz high pass.