Cochlear Place Specificity of the Auditory Brainstem Response to Narrowband Chirp Versus 2-1-2 Stimuli: High-Pass Noise/Derived Responses.

OBJECTIVES: In recent years, many researchers have recommended using narrowband chirp (NBchirp) stimuli for auditory brainstem response (ABR) audiometry instead of more-standard 2-1-2 cycles linear-gated tones, primarily because NBchirps often result in larger ABR wave V amplitudes. However, the acoustic frequency spectra of currently recommended NBchirps are wider than those for 2-1-2 tones, and it is currently not known whether ABRs to these NBchirps have similar (or poorer) cochlear place specificity compared with 2-1-2 tones. The present study used the high-pass noise/derived response technique to assess the cochlear regions contributing to ABRs evoked by NBchirp versus 2-1-2 stimuli. DESIGN: A total of 24 adults with normal hearing participated (N = 12 for each stimulus frequency). Stimuli were 60-dB pe SPL 500- and 2000-Hz NBchirps and 2-1-2 linear-gated tones mixed with high-pass (HP) filtered masking noise. The level of broadband (pink) noise required to mask the ABR was determined individually; then the broadband noise at this level was HP-filtered at ½-octave intervals. At least 3 ABR replications were obtained for each condition, with recordings stopped when the residual noise level of each replication was reduced to about 40 nV, with the goal of the average of all replications having a residual noise of about 23 nV (for HP noise-masked conditions). Derived responses (DRs) representing 1-octave-wide or ½-octave-wide cochlear regions were calculated by subtracting ABRs recorded in HP noise. RESULTS: Non-masked ABR amplitudes in response to NBchirps were significantly larger than those to 2-1-2 stimuli, averaging 55% larger for 500 Hz and 81% larger for 2000 Hz. For both 500- and 2000-Hz stimuli, HP noise masking produced significant amplitude decreases, occurring 1 to ½ octave higher for ABRs to NBchirps versus 2-1-2 tones. One-octave-wide and ½-octave-wide DR amplitude profiles for the ABRs to 2-1-2 tones showed good cochlear place specificity, as described in previous studies. DR results for the NBchirps were similar but showed important differences. The profiles for the 2000-Hz NBchirps showed significantly larger amplitudes in the 4- and 1-kHz DRs compared with the 2-1-2 stimuli. Many more responses were seen 1 octave away for the 2000-Hz NBchirp compared with the 2-1-2 tone. DR results for 500-Hz tones showed similar patterns, but differences did not quite reach statistical significance, except that amplitudes to NBchirps were larger at DR354, DR500, and DR707. A measure of the width of the 1-octave-wide and ½-octave-wide DR amplitude profiles (BW0.075, in Hz) showed that the 500- and 2000-Hz NBchirp profiles were significantly wider (32 to 77%) compared with those for 2-1-2 stimuli. As the cochlear area able to respond decreased, wave V amplitudes to NBchirp stimuli decreased more than those for 2-1-2 stimuli, with smaller differences between stimuli for ½-octave-wide responses. Objective estimates of response signal to noise ratio supported the wave V amplitude results, with some differences. CONCLUSIONS: ABRs to NBchirps reflect wider cochlear contributions than those to 2-1-2 tones. Responses to NBchirps arise from cochlear regions as far as one octave away from the stimulus frequency. In contrast, responses to 2-1-2 tones arise from cochlear regions primarily within approximately ±0.5 octaves of the stimulus frequency. Further research in individuals with hearing loss, especially with steep hearing loss, is required to determine whether the wider bandwidths for NBchirps result in threshold misestimation and whether NBchirp amplitude advantages over more-standard stimuli remain with hearing loss.