Simultaneous recording of derived ABR and click-evoked otoacoustic emissions as estimates of cochlear frequency specificity

Derived auditory brainstem response (dABR) wave V latencies (V$\sb {\rm d})$ from narrow frequency regions are interpreted as reflecting the cochlear travel time to that frequency region in addition to a small neural delay. Click evoked otoacoustic emission (CEOAEs) latencies are believed to reflect the cochlear travel time to that frequency region and back to the earcanal. ABR and CEOAEs were recorded simultaneously for clicks presented at 45 dB SL with four different high-pass noise maskers (1, 2, 4 and 8 kHz cutoff frequencies) in eight normal hearing adults. In an attempt to achieve noise cancellation in the CEOAE recordings, a 333 msec sample of random noise was alternated with the same sample of noise inverted in phase. The signals were then averaged together. Incomplete noise cancellation, due to noise-related evoked otoacoustic emissions, precluded reliable estimates of CEOAE latencies under the different noise conditions. Consequently, latencies for corresponding frequency bands were obtained by band-pass filtering the unmasked CEOAEs between 4-8 kHz, 2-4 kHz, 1-2 kHz and 0.3-1 kHz. The change in latency with frequency for both the CEOAE and the dABR was consistent with previously published data. Relative latency shifts between adjacent frequency bands of the CEOAE were approximately twice those of the dABR shifts reflecting the need for the CEOAE to travel to the frequency region and back towards the middle ear. Two pieces of evidence suggest that the addition of the high-pass maskers had an effect on the pattern of activity near the apex. (1) The latency of the dABR for the 8 kHz high-pass noise (when detectable) was always shorter than the unmasked ABR. (2) The spectral characteristics of CEOAE responses in the 1-2 kHz region were significantly modified by the addition of 4 kHz high-pass noise. Despite this evidence that the high-pass noise does modify cochlear response, the similarities between the estimates of travel time from the dABR and the unmasked CEOAE suggest that the latter may still be useful as a clinical tool to evaluate frequency specificity