POPULATION STUDY OF COCHLEAR NERVE-FIBERS - COMPARISON OF SPATIAL DISTRIBUTIONS OF AVERAGE-RATE AND PHASE-LOCKING MEASURES OF RESPONSES TO SINGLE TONES

Spontaneous rates (SR) of cochlear nerve fibers clearly show a bimodal distribution. This was seen in large samples obtained from individual cats, as well as for a pooled sample of 1,443 nerve fibers from 12 cats. The boundary region between the low-SR and the high-SR populations is 10-20 spikes/s. The low-SR population with SR<15 spikes/s comprises 24% and the high-SR population with SR≥15 spikes/s comprises 76% of the total sample. The mode of the high-SR population in the pooled sample is 50-60 spikes/s. Spatial distributions of cochlear nerve fiber responses have been studied by recording from hundreds of cochlear nerve fibers in individual animals and by using Fourier analysis to obtain the amplitude R(i) and the phase θ(i) of the ith harmonic component from each of the period histograms. The spatial distributions of (R0-SR), (R0/SR), R1, (R1/SR), and (R1/R0) of responses to a 1,000-Hz single tone at a sound pressure level of 20 dB (re 20 μN/m2) all show a global maximum amplitude at the position where the characteristic frequency (CF) is equal to the stimulus frequency. In addition, a secondary peak of R1 has been observed in the CF region of 7-8 kHz. As the sound pressure level increases to 45 and 70 dB, the spatial distributions of the response measures become complex, showing more peaks. The spatial distributions of the response measures R2/SR and R3/SR (the second and the third harmonic distortion components divided by SR) of responses to 1,000-Hz single tones at sound pressure levels of 20-70 dB (re 20 μN/m2) are qualitatively similar to the distribution of the response measure R1/SR in that the spatial positions of the visible peaks of all three are the same for a given sound pressure level. This suggests that these single-tone harmonic distortion components are locally generated at the position of observation, rather than propagated from a remote place in the cochlea. Results from cochlear nerve response-mapping experiments using single tones of 1,000 Hz and other frequencies ranging from 300 to 3,600 Hz show: a) that separation of cochlear nerve fibers into low-SR and high-SR subpopulations facilitates visualization of spatial patterns of neural responses; b) that R0 does not show a clear spatial pattern of neural responses, whereas each of (R0-SR), R0/SR, R1, R1/SR, R1/R0, R2/SR, and R3/SR does; c) that phase-locking response measures R1, R1/SR, and R1/R0 sensitively reflect excitation amplitude even when the excitation amplitude is too small to cause a measurable increase in R0; and d) that response measures derived from R2 and R3 add relatively little information to that derived from R1 concerning spatial patterns of neural responses.