Limited segregation of different types of sound localization information among classes of units in the inferior colliculus

Abstract

The auditory system uses three cues to decode sound location: interaural time differences (ITDs), interaural level differences (ILDs), and spectral notches (SNs). Initial processing of these cues is done in separate brainstem nuclei, with ITDs in the medial superior olive, ILDs in the lateral superior olive, and SNs in the dorsal cochlear nucleus. This work addresses the nature of the convergence of localization information in the central nucleus of the inferior colliculus (ICC). Ramachandran et al. (1999) argued that ICC neurons of types V, I, and O, respectively, receive their predominant inputs from ITD-, ILD-, and SN-sensitive brainstem nuclei, suggesting that these ICC response types should be differentially sensitive to localization cues. Here, single-unit responses to simultaneous manipulation of pairs of localization cues were recorded, and the mutual information between discharge rate and individual cues was quantified. Although rate responses to cue variation were generally consistent with those expected from the hypothesized anatomical connections, the differences in information were not as large as expected. Type I units provide the most information, especially about SNs in the physiologically useful range. Type I and O units provide information about ILDs, even at low frequencies at which actual ILDs are very small. ITD information is provided by a subset of all low-frequency neurons. Type V neurons provide information mainly about ITDs and the average binaural intensity. These results are the first to quantify the relative representation of cues in terms of information and suggest a variety of degrees of cue integration in the ICC.

Figure 1.

Response maps typical of the classes of ICC units exhibiting sustained responses to tones. Driven rates are calculated over a 200 ms tone duration, and spontaneous rates are subtracted to reveal inhibition. At 0 dB attenuation (attn), the sound level is ∼100 dB sound pressure level. A, Type V unit, characterized by EE responses and little to no inhibition. B, Type I unit, characterized by EI responses and sideband inhibition. C, Type O unit, characterized by an island of excitation at low levels of contralateral stimulation turning to inhibition at higher levels. Responses to ipsilateral tones are typically inhibitory.

Figure 2.

Stimulus construction. A, ITD/ILD stimulus set. Frozen noise was filtered through a cat HRTF, averaged over spatial location to give the location-independent spectral characteristics of the pinnae and ear canal. The noise was then copied into two streams, on which an ITD and an ILD were placed. B, SN/ILD stimulus set. The frozen noise was filtered through one of five HRTFs showing a prominent SN, split into two streams and attenuated relative to one another. C, ABI/ILD stimulus set. As in A, the frozen noise was filtered through the spatially averaged HRTF. It was then attenuated to one of five levels before being split into two streams that are attenuated relative to one another. The ILD was imposed such that the ABI was preserved. D, The five HRTFs used in the SN/ILD stimulus set. All were taken from the midline, at 0° azimuth. Note the prominent first SN, which shifts to higher frequencies as the elevation is increased. The SN/ILD stimulus set was resampled before presentation such that the notch for 15° elevation fell on the BF. avg, Average; attn, attenuation; contra, contralateral; ipsi, ipsilateral; DC, zero frequency.

Figure 3.

Extracting the MI to the full stimulus set and to the individual location cues. A, The surface shows the mean response rate as a function of the ILD and SN (surface-smoothed with cubic spline interpolation). Contours on the vertical walls of the plot show the mean (±1 SD) of the rate as a function of ILD or SN only. The debiased MI is given above the plot with the bias estimate in parentheses. B, Mean rate versus ILD. C, Mean rate versus SN. D-F, Same for another unit showing more SN sensitivity. sp, Spikes.

Figure 4.

Rate profiles of responses to the SN/ILD and ITD/ILD stimulus sets. Each profile has been normalized to its maximum response. Thin gray lines represent individual rate profiles, and thick black lines represent the means of all rate profiles in each category. oct, Octave.

Figure 5.

MI plotted against the BF for the SN/ILD and ITD/ILD stimulus sets. SN/ILD stimuli are not differentiated from ITD/ILD stimuli, as the two groups overlap. A, MI_FULL. B, MI_ILD. C, MI_SN. D, MI_ITD. Freq, Frequency.

Figure 6.

The proportion of MI_FULL used in coding individual cues. A, MI_SN is plotted against MI_ILD for all units studied with the SN/ILD stimulus set, both expressed as a fraction of MI_FULL. B, Same for MI_ITD versus MI_ILD. Only units with BFs <4 kHz are plotted in B.

Figure 7.

Normalized rate profiles of responses to the ABI/ILD stimulus set. Plotting conventions are the same as for Figure 4.

Figure 8.

The differential coding of the average level and level difference information. A, MI_ILD versus BF for the ABI/ILD stimulus set. B, MI_ABI versus BF. C, MI_ABI plotted against MI_ILD for each unit, expressed as a percentage of MI_FULL. Freq, Frequency.

Figure 9.

Binaural sensitivity and coding. A, Histogram of monaural indices broken down by unit type. Values near 0 are the most binaural. Type O units cluster near 0, whereas type V units are more broadly distributed. B, MI_ILD is plotted against the absolute value of the monaural index, shown on a log scale. C, MI_ABI versus the monaural index. Mon Ind, Monaural index.