The effect of high-pass noise on the scalp-recorded frequency following response (FFR) in humans and cats

Abstract

Low-frequency tone-burst stimuli were used to elicit the scalp-recorded frequency following response (FFR) from normal-hearing adult human subjects and anesthetized adult cats. High-pass masking stimuli (with variable low-frequency cutoffs) were presented simultaneously with these tone-burst stimuli. For a small group of human subjects, analogous psychophysical masking curves (threshold shift and binaural loudness balance) were conducted so that physiologically derived curves (masked FFR) could be related to these psychophysically derived curves. Among the results were the following: (1) human and cat FFR masking curves were very similar; (2) at moderate to high probe-tone stimulus levels, the amplitude of the FFR masked with a 2-kHz cutoff was approximately 70% to 75% of the amplitude of the unmasked FFR; (3) at probe-tone stimulus levels near FFR threshold, the amplitude of the FFR masked with cutoffs of 1.5, 1.8, and 2.0 kHz was equivalent to the amplitude of the unmasked FFR; (4) a sizable fraction (10% to 30%) of the masked FFR amplitude remained during masking by stimuli with low cutoffs (about 900 Hz or less); (5) the latency of the FFR recorded from humans and cats (using probe-tone levels of 80-dB peak SPL or greater) increased systematically at lower and lower masker cutoffs; (6) at lower probe-tone levels, FFR latency recorded from cats did not increase using masker cutoffs of 2.0, 1.8, and 1.5 kHz, but with the use of lower-frequency cutoffs, latency values obtained from the cat data showed systematic increases similar to those previously cited for humans; and (7) FFR masking curves did not correspond with psychophysically derived curves except at low probe-tone stimulus intensities. These results suggest that although the area of stimulation (along the basilar membrane) at stimulus frequencies and intensities at which the FFR is commonly generated is broad, the most prominent input arises from the apical half and third of the cochlea (below 2 kHz) for man and cat, respectively. Except for the lower stimulus levels employed in the animal experiments, the responding area was apparently not restricted to a narrow band around the 500-Hz signal. These results are consistent with the interpretation that as the probe-tone level is raised, a greater percentage of FFR amplitude will be accounted for by neurons tuned to high frequencies. Also, as the masking signal level is raised, these results are consistent with the interpretation that complications inherent in simultaneous masking procedures (specifically, masking by low-frequency components in the acoustic stimulus and remote masking by combination components generated by the masker) do not allow masking paradigms to provide a precise definition of the cochlear sites of FFR generation along the basilar membrane.