Effects of interaural time delays of noise stimuli on low-frequency cells in the cat's inferior colliculus. I. Responses to wideband noise

1. We examined the responses of low-frequency neurons in the central nucleus of the inferior colliculus (ICC) of the cat to interaurally delayed, wideband noise stimuli. The stimuli were pseudorandom noise signals that were generated digitally with a nominal bandwidth of 60-4,000 Hz. We also compared the responses to noise with those obtained from interaural phase differences of pure tones. 2. We studied 144 neurons with characteristic frequencies below 2.5 kHz. Eighty-five percent of these were sensitive to changes in both interaural time differences (ITDs) of noise and interaural phase differences of pure tones, only 2% were sensitive to one stimulus but not the other, and the remainder were insensitive to both stimuli. 3. For most cells the discharge rate was modulated in an approximately cyclic fashion by changes in ITDs of the wideband noise stimuli. The maximal spike counts often occurred near zero ITD, and there was considerable variability in the nature of the cycling, though it usually disappeared for ITDs greater than ±4,000 μs. The position of the central peak was usually (65%) within the physiologically relevant range of ±400 μs, and most (80%) occurred at positive ITDs, which corresponded to delays to the ipsilateral stimulus. In general, the shapes of the responses were not affected by changes in stimulus level above threshold. 4. As long as identical noises were delivered to both ears, the responses were not sensitive to the particular noise stimulus used. When uncorrelated noises were delivered to the two ears, there was no sensitivity to ITDs. 5. Composite curves were computed by linear summation of the responses to ITDs of pure tones at frequencies spaced at equal intervals throughout each cell's response area. The shapes of composite curves were similar to the responses of the same cell to ITDs of wideband noise stimuli. The positions of the central peaks of these two functions were highly correlated (r = 0.91, slope = 0.97). 6. The values of characteristic delay and characteristic phase computed from the tonal responses were found to be good indicators of the shapes of the noise delay curves. Characteristic phases (CPs) near zero were associated with noise delay curves symmetric about the central peak, CPs near 0.5 cycles with those symmetric about the trough, while CPs between 0 and 0.5 or between 0.5 and 1.0 had noise delay curves that were asymmetric with a prominent trough to the left or right, respectively, of the central peak. 7. The temporal characteristics of the responses to interaurally delayed noise were found to be characterized best by the synchronized-rate curve, which was computed by taking the product of spike count and the synchronization coefficient of the interaural phase sensitivity curve. The median frequency (MF) of the sync-rate curve was highly correlated with the response frequency (i.e., the reciprocal of the period between adjacent peaks) of the noise delay curve (r = 0.90, slope = 1.05). Our preliminary results indicate that the sync-rate and noise delay curves may be considered to be Fourier transform pairs. 8. There was a positive and significant correlation between the reciprocal of the temporal tuning width (TT) of the responses to interaurally delayed noise and the MF of the sync-rate curve (r = 0.56, slope = 0.002). When the regression line between 1/TT and the sync-rate MF was extrapolated out to 8 kHz, the data available from the inferior colliculus (IC) of the barn owl fell close to this function, suggesting a similar neural mechanism for binaural interaction in the cat and barn owl. 9. Our results indicate strong support for a cross-correlation model for interaural time sensitivity of low-frequency neurons in the ICC. Furthermore, the responses to interaural delays of sideband noise stimuli could be predicted from linear summation of the responses to interaural phase differences of tones equally spaced within the cell's response area.