| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 79 No. 1 January 1998, pp. 253-269
Copyright ©1998 by the American Physiological Society
1 Department of Neurophysiology, University of Wisconsin Medical School, Madison, Wisconsin 53706; and 2 Division of Neurophysiology, Medical School, University of Leuven, B-3000 Leuven, Belgium
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Joris, Philip X. and Tom C. T. Yin. Envelope coding in the lateral superior olive. III. Comparison with afferent pathways. J. Neurophysiol. 79: 253-269, 1998. Binaural cues for spatial localization of complex high-frequency sounds are interaural level and time differences (ILDs and ITDs). We previously showed that cells in the lateral superior olive (LSO) are sensitive to ITDs in the envelope of sinusoidally amplitude-modulated (AM) signals up to a modulation frequency of only ~800 Hz. To understand the limitations in this ITD-sensitivity, we here compare responses to monaural modulation in LSO and its input pathways, derived from cochlear nucleus and medial nucleus of the trapezoid body. These pathways have marked functional and morphological specializations, suggestive of adaptations for timing. Afferent cell populations were identified on the basis of electrophysiological signatures, and for each population, average firing rate and synchronization to AM tones were compared with auditory-nerve fibers and LSO cells. Except for an increase in modulation gain in some subpopulations, synchronization of LSO afferents was very similar to that in auditory nerve fibers in its dependency on sound pressure level (SPL), modulation depth, and modulation frequency. Distributions of cutoff frequencies of modulation transfer functions were largely coextensive with the distribution in auditory nerve. Group delays, measured from the phase of the response modulation as a function of modulation frequency, showed an orderly dependence on characteristic frequency and cell type and little dependence on SPL. Similar responses were obtained to a modulated broadband carrier. Compared with their afferents, LSO cells synchronized to monaurally modulated stimuli with a higher gain but often over a narrower range of modulation frequencies. Considering the scatter in afferent and LSO cell populations, ipsi- and contralateral responses were well matched in cutoff frequency and magnitude of delays. In contrast to their afferents, LSO cells show a decrease in average firing rate at high modulation frequencies. We conclude that the restricted modulation frequency range over which LSO cells show ITD-sensitivity does not result from loss of envelope information along the afferent pathway but is due to convergence or postsynaptic effects at the level of the LSO. The faithful transmission of envelope phase-locking in LSO afferents is consistent with their physiological and morphological adaptations, but these adaptations are not commensurate with the rather small effects of physiological ITDs reported previously, especially when compared with effects of ILDs. We suggest that these adaptations have evolved to allow a comparison of instantaneous amplitude fluctuations at the two ears rather than to extract interaural timing information per se.
The auditory nerve (AN) provides the CNS with ample timing information about the envelope of amplitude-modulated (AM) signals (Cooper et al. 1993
Most of our methods are as described in previous papers (Joris 1996
Data were obtained for 653 single units (not including AN fibers), of which we selected 14 SBCs, 79 GBCs, 37 MNTB, and 24 LSO cells according to the physiological criteria stated in METHODS. MNTB and LSO cells also were localized histologically to the appropriate region, except for four MNTB cells from one animal, for which histology was not available, and five presumed LSO recordings that were localized rostral to LSO (see Joris 1996 Changing SPL
In the AN, envelope synchronization is nonmonotonically dependent on SPL (Cooper et al. 1993
Changing modulation frequency
SYNCHRONIZATION MAGNITUDE.
Envelope ITD-sensitivity requires temporal envelope information at the site of binaural interaction. This information, supplied to the CNS via the AN, deteriorates at high modulation frequencies: modulation transfer functions, which are graphs of synchronization as a function of modulation frequency, are low-pass (Joris and Yin 1992
SYNCHRONIZATION PHASE.
As in AN fibers, LSO cells and their afferents show a linear accumulation of phase lag with modulation frequency. The examples in Fig. 10 show that the slope of the cumulated phase-frequency functions is related to cell type, being largest for LSO cells and shortest for AN fibers. Moreover, the slopes for ipsilateral and contralateral modulation in LSO are well matched, but the functions are offset by ~0.5 cycle due to the different sign of the input (excitatory or inhibitory) (Joris 1996
AVERAGE RATE.
With increases of modulation frequency >300 Hz, ITDs are decreasingly effective in modulating the firing rate of LSO cells (Joris 1996
ADDITIONAL FINDINGS.
The data presented so far give a coherent picture of envelope phase-locking in the LSO circuit: LSO afferents show strong phase-locking over a wide range of modulation frequencies; this is translated by LSO cells into ITD-sensitivity over a more limited frequency range. To put these findings in perspective, we report some additional observations, for smaller samples of cells, with regard to the effect of SPL on modulation transfer functions, the envelope phase-locking when a broadband carrier is used, and finally how bushy cells compare with the other major AVCN cell type: the chopper/stellate cells.
Two generalizations can be made from the current results. 1) There is a high degree of conservation in the magnitude of envelope phase-locking in both ipsi- and contralateral afferent channels of the LSO circuit for all stimulus parameters, which ensures that the LSO is supplied with envelope information over the widest possible range of modulation frequencies and well matched for the two sides in terms of amplitude and phase. 2) Because cells interposed between the AN and LSO have AM responses that resemble the AN rather than the LSO, the previously noted differences between envelope phase-locking in LSO and AN reflect input convergence or intrinsic properties at the level of LSO.
Limits in LSO afferents
All LSO afferents showed low-pass modulation transfer functions very similar to those of AN fibers (Figs. 5 and 9), with 3-dB cutoff frequencies that largely overlapped with the AN distribution (Figs. 7 and 12B). Only at CFs > 20 kHz, where measurements were sparse, were GBC and MNTB cutoffs in the lower part of the AN range. Dependence of synchronization on SPL in LSO afferents was nonmonotonic as in the AN (Fig. 2). Maximum synchronization values (Fig. 4) of GBCs and MNTB cells were higher than AN fibers at CFs below ~7 kHz but covered the same range at higher CFs. Neither in synchronization magnitude nor3-dB cutoff frequency range was there any indication of a difference between SBCs and AN fibers (Figs. 4, 7, and 12) although our sample of SBCs >10 kHz is very limited.
Monaural and binaural limits in LSO cells
In contrast to the general similarities in envelope synchronization between AN and LSO afferents, responses of cells in LSO to monaural modulation deviated in several ways from their afferents, including higher maximum synchronization values (Fig. 4) and a reduced modulation frequency range (Fig. 7). Phenomenologically, these transformations parallel findings on pure tone synchronization in CN bushy cells, which also is enhanced in magnitude but restricted in frequency range relative to the AN (Joris et al. 1994a MAXIMUM SYNCHRONIZATION.
High synchronization values of LSO or presumed LSO cells to ipsilateral modulation were reported earlier (Batra et al. 1997b PHASE-LOCKING FREQUENCY RANGE.
Synchronization of LSOcells to the envelope difference frequency in a binaural AM beat stimulus of increasing modulation frequency is restricted to modulation frequencies AVERAGE RATE.
Average firing rate of LSO cells decreased in most cells for increasing frequency of either contra- or ipsilateral modulation, as we found previously for binaural modulation (Fig. 13). Because this decrease was stronger or even in the opposite direction than would be predicted from average rate changes in afferents, it must be caused by temporal factors in the binaural interaction. The rate decrease for contralateral modulation is presumably due to loss of envelope phase-locking at high modulation frequencies. With increasing modulation frequency, the MNTB input to LSO becomes effectively demodulated or sustained. As discussed previously (Joris 1996 Interaural timing
Previous studies of LSO have shown that ipsi- and contralateral responses are well matched for a variety of tonal suprathreshold response measures (Boudreau and Tsuchitani 1970
Cell classification
We chose physiological criteria that optimized sorting of cells to the appropriate categories, but this procedure likely also biased our sample toward cells with pronounced characteristics. For example, one of our requirements for classification of a cell as GBC was a clear PLN response pattern, which may have biased the GBC sample toward cells with good timing properties. Indeed, a small number of GBCs identified anatomically by their calyceal ending in the MNTB maintain a PL response pattern at high SPLs (Smith et al. 1991 Relevance of binaural envelope information
We conclude this series of papers by posing a paradox. The afferent pathways to LSO are characterized by the presence of specializations in membrane properties, axon diameters, and size and placement of terminals. These features minimize differences between total ipsi- and contralateral conduction delay and limit spatial and temporal integration. Speculation on their purpose traditionally has been in terms of timing, which motivated the present series of studies of ITD-sensitivity in LSO. Humans have a well-documented ability to lateralize high-frequency sounds on the basis of envelope ITDs. The envelope ITD-sensitivity that we described in the preceding papers showed a number of parallels with human performance. Psychophysically, the smallest ITDs that can be discriminated approach those for low-frequency fine-structure (Henning 1974 LSO as a multiplexer
This teleological argument by itself does not specify the benefit of maintaining the envelope in the extraction of ILD. That temporal and spectral accuracy are so well preserved at the input stage suggests that the LSO might have a broader role in spatial hearing than extracting ILD information for coding of azimuthal position of a single, static source. A first role may be temporal multiplexing of ILDs. By virtue of the preservation of temporal envelope information in the afferent inputs, LSO cells are able to encode quick alterations in ILD, e.g., as generated by multiple, modulated sound sources with overlapping spectra but differing spatial location. It is important to note here that the ILD coding in LSO is constrained by the absence of tuning for this cue. Cells do not show a characteristic or best ILD by which the cue value would be mapped as a focus of activity in a cell population graded for such property. Rather, the available evidence supports a scalar coding of ILD
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Javel 1980; Joris and Yin 1992; Palmer 1982; Wang and Sachs 1993). Following early studies by Møller (1974), there has been a renewed interest in temporal envelope coding at the first level of synaptic integration: the cochlear nucleus (CN) (Frisina et al. 1990; Rhode and Greenberg 1994; Wang and Sachs 1994). Different cell types in the CN transform the envelope information supplied by the AN in different ways, but it is unclear whether and how this information is used behaviorally. The most convincing case for the use of envelope cues, as for temporal cues in general, comes from binaural studies. We describe the sequential transformations of envelope information in an anatomically and physiologically well-characterized binaural circuit, in which we previously demonstrated sensitivity to envelope interaural time differences (ITDs) of high-frequency AM signals (Joris and Yin 1995).
; Roth et al. 1980). Human subjects can detect ITDs of the carrier at very small values (<20 µs), but only at frequencies below ~1.5 kHz. ITDs of the envelope are detected most easily for high-frequency carriers (Henning 1974). However, acoustically, physiologically, and psychophysically, ILDs are the most prominent localization cue at high frequencies. Physiological sensitivity to ITDs of high-frequency AM stimuli was first described in the inferior colliculus (Batra et al. 1989, 1993; Yin et al. 1984).
View larger version (17K):
[in a new window]
FIG. 1.
Cartoon of the lateral superior olive (LSO) circuit and the physiological signatures used to identify afferent responses. Recordings from excitatory spherical bushy cells (SBCs) in the anteroventral cochlear nucleus (AVCN) show a primary-like response to short tonebursts at their characteristic frequency (CF) and a prepotential in the extracellular spike waveform. Recordings from axons of excitatory globular bushy cells (GBCs) in the ipsilateral trapezoid body (TB) show a primary-like-with-notch pattern to short (25 ms) tonebursts. Inhibitory cells in the medial nucleus of the trapezoid body (MNTB) also show a prepotential and a primary-like-with-notch pattern as in GBCs to stimulation of the contralateral ear. AN, auditory nerve.
; Joris and Yin 1992; Palmer 1982). This limit could be imposed by the afferents. Second, the LSO circuit has marked functional and morphological specializations in both of its limbs, suggestive of adaptations for timing. The three cell types involved in the transmission of signals from auditory nerve to LSO all have a "bushy" morphology, i.e., sparse and short dendrites (Brawer et al. 1974), and derive most of their input from a limited number of afferents through large axosomatic endings. AN fibers give large terminals to bushy cells: a single or few (2-5) so-called endbulbs of Held to SBCs (Melcher 1993; Ryugo and Sento 1991), and modified endbulbs from an estimated 10-25 auditory nerve fibers to GBCs (Liberman 1991; Spirou et al. 1990). The GBC input to MNTB cells is in the form of a single large calyx of Held (Smith et al. 1991; Spirou et al. 1990; Tolbert et al. 1982), apparently the first described and largest terminal in the mammalian brain (Jean-Baptiste and Morest 1975). SBCs give excitatory endings on the distal dendrites of ipsilateral LSO cells, whereas the inhibitory input to LSO from MNTB is placed on the proximal dendrites and soma (Cant 1984; Glendenning et al. 1991; Helfert et al. 1989; Zook and DiCaprio 1988; P. H. Smith, P. X. Joris, and T.C.T. Yin, unpublished results). Axon diameters of contralateral afferents are larger than those of ipsilateral afferents (see DISCUSSION). The bushy cells in AVCN and MNTB also share membrane properties that allow for a precise transmission of the temporal information available in their afferents (Banks and Smith 1992; Banks et al. 1993; Borst et al. 1995; Forsythe and Barnes-Davis 1993; Manis and Marx 1991; Oertel 1983; Smith and Rhode 1987; Wu and Kelly 1993; Wu and Oertel 1984).
). Similarly, responses of MNTB principal cells are nearly identical to those of GBCs (Guinan et al. 1972a,b; P. H. Smith, P. X. Joris, and T.C.T. Yin, unpublished results). Despite the longer path length and extra synapse in the contralateral branch of the LSO circuit, the latencies of LSO responses to ipsi- and contralateral stimulation appear similar (Boudreau and Tsuchitani 1968; Tsuchitani 1988).
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
; Joris and Yin 1995). We obtained extracellular recordings from cats anesthetized with pentobarbital. The surgical approach and recording techniques were identical to those in our previous reports, and LSO and afferent responses often were obtained within the same penetration. Responses can be recognized as being derived from LSO afferents with the aid of various electrophysiological signatures (Bourk 1976; Guinan et al. 1972a,b; Smith et al. 1991; Spirou et al. 1990), examples of which are shown in Fig. 1. Cells in AVCN with spikes preceded by prepotentials (PPs) and with primary-like (PL) responses to tones at characteristic frequency (CF) were considered SBCs. Primary-like-with-notch (PLN) responses recorded in the trapezoid body ipsilateral to the excitatory ear were considered to be from GBC axons. PLN responses, likely also from GBCs, also were encountered in the contralateral trapezoid body but are not included here because of the possible confusion with MNTB responses. PP units in the ventral brain stem, driven exclusively by the contralateral ear, were considered MNTB cells. Spikes were monitored on-line for PPs with an averaging oscilloscope. When a PP or complex waveform was present, the spike complex was averaged at a sampling rate of 40 kHz and a bandwidth of 0.1-10 kHz for 
200 occurrences. In the vast majority of cases, the PP was also visible in the ongoing neural signal without averaging. LSO cells were identified histologically and on the basis of binaural contralateral inhibition, ipsilateral excitation (IE)-sensitivity and chopper responses to ipsilateral tone bursts at CF as described in Joris and Yin (1995). The "LSO-EE" cells described by Joris (1996) are not considered here. We also recorded AM data from monaural AVCN cells that were classified as choppers based on the multimodal poststimulus time histogram (PSTH) in response to short CF tone bursts as well as on response irregularity, quantified by the coefficient of variation (CV) calculated over the full range of SPLs (Young et al. 1988). Finally, we make use of 186 AN fibers from a previous study (Joris and Yin 1992) and 9 additional AN fibers (see Fig. 15) recorded using the same methods.
View larger version (17K):
[in a new window]
FIG. 15.
Comparison of 3 measures of envelope synchronization with wideband (ordinate) and tonal (abscissa) carriers. Each symbol represents data for 1 cell or axon, and different symbols are used for different afferent categories (inset). Three categories of cells did not fullfil all the criteria stated in METHODS and were classified on the basis of poststimulus time histogram: PHL, PL, and PLN refer to cells with phase-locked, primary-like, or primary-like-with-notch response to carrier, respectively. A: maximum synchronization values, derived from a level series. B: 3-dB cutoff frequency of modulation transfer functions. C: delay measured from phase-frequency functions. - - -, diagonals of equality. Slopes and y intercepts of linear regressions were A: 0.96 and 
0.016, B: 0.96 and 60 Hz, and C: 1.01 and 0.03 ms.
) and LSO (Joris 1996; Joris and Yin 1995). The general approach was to optimize stimulus parameters for maximal monaural and/or binaural phase-locking in each cell, while still keeping a high average spike rate. AM stimuli usually were modulated fully (modulation depth m = 100%), and the carrier frequency always was placed at the frequency to which the cell was most sensitive (= CF). Duration was 600 ms, repeated every second for typically 20 or 40 times. Magnitude and phase of synchronization at the envelope frequency were quantified with the vector strength R, tested for significance (P < 0.001) with the Rayleigh test and expressed as a gain measure defined by 20 log (2R/m).
35 dB). Binaural sensitivity of LSO cells was assessed with a binaural AM beat stimulus, typically consisting of a 5-s-long AM stimulus delivered with a 1-Hz difference in modulation frequency to the two ears.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
). To complement our small sample of SBCs, we report on 47 units that did not fulfill all the SBC criteria: 6 PP units recorded in AVCN but with undetermined PSTH, 17 PL units recorded in AVCN but lacking a PP, and 24 PL units recorded in the trapezoid body (TB). This sample may be "contaminated" with GBCs, which can show PL responses (Smith et al. 1991) and PPs (Bourk 1976). For comparative purposes, AM data also were obtained in 20 AVCN choppers.
; Joris and Yin 1992; Smith and Brachman 1980; Wang and Sachs 1993), and the same was true in LSO afferents. Figure 2 shows average rate (*, left ordinate) and synchronization at the modulation frequency (
, right ordinate) as a function of SPL for one cell of each afferent class. In every case, the rate-level function is sigmoidal, and the synchronization-level function nonmonotonic. R reaches a maximum slightly above rate threshold, on average (in dB re. tuning curve threshold, ± SD) at 2.4 ± 6.6, 6.0 ± 7.0, and 5.5 ± 4.8 for SBCs, GBCs, and MNTB cells, respectively. The average in AN is 10.0 ± 4.95 (Joris and Yin 1992).
View larger version (25K):
[in a new window]
FIG. 2.
Synchronization-level functions for an afferent fiber of each category. Average rate (left ordinate, *) showed a sigmoidal, monotonic increase in all fibers. Synchronization at the envelope frequency (right ordinate, ) was always nonmonotonic. 
, nonsignificant synchronization (P < 0.001). CFs were as indicated, modulation frequency was 100 Hz.
View larger version (16K):
[in a new window]
FIG. 3.
Synchronization-level functions in a LSO cell to ipsilateral (left) and contralateral (right) modulation. Abscissa indicates the sound pressure level (SPL) of the modulated waveform: the stimulus at the other ear was unmodulated and held at 44 dB. Modulation frequency was 100 Hz. Only significant synchronization values are graphed.
; Kim et al. 1990; Rhode and Greenberg 1994; Wang and Sachs 1994). We therefore compared the slope of a linear regression through this part of the function with measurements previously obtained in the AN with the same procedure (Joris and Yin 1992). These slopes were very similar in AN fibers, bushy cells, and MNTB cells [sample size (dB
1); AN: 186, 0.012 ± 0.003; SBC: 9, 0.010 ± 0.0013; GBC: 41, 0.011 ± 0.002; MNTB: 21, 0.011 ± 0.002]. However, the slopes for ipsilateral LSO responses (n = 21, 0.007 ± 0.0039) were significantly more shallow than in the AN (P < 107, Mann-Whitney U), indicating less of a decrease in R at high SPLs.
; Joris et al. 1994a,b) and also have been reported to have improved synchronization to envelopes (Frisina et al. 1990; Joris et al. 1994a; Rhode and Greenberg 1994; Wang and Sachs 1994). For each fiber, we obtained the value of maximal significant synchronization from the synchronization-level function. These values are shown at each cell's CF in the scatter diagram of Fig. 4 (see legend for means and SD). A first striking feature is the general coextensiveness of AN and bushy cell values. For the small sample of SBCs, this seems to be the case over the entire CF range that was sampled. GBCs and MNTB cells diverge from AN fibers below ~7 kHz because of opposite tendencies: whereas AN fibers tend to show lower synchronization values at low CFs (Joris and Yin 1992; Wang and Sachs 1993), GBCs and MNTB cells tend to show higher values at low CFs than at high CFs. A second feature is that LSO cells typically reach higher maximal R values than their afferents, at all CFs and for both ipsi- and contralateral monaural modulation. As reported earlier (Joris and Yin 1995) and as expected for an IE-type interaction (Joris 1996), R values to binaural stimulation are high (Fig. 4; 
) when the envelopes to the two ears are out of phase.
View larger version (26K):
[in a new window]
FIG. 4.
Maximum synchronization values (Rm) for different cell classes, derived from synchronization-level functions. For LSO afferents, 1 datapoint is shown per cell. Up to 3 kinds of measure are shown for LSO cells: maximal synchronization to monaural ipsi- or contralateral modulation (filled triangles), and to binaural modulation ( ). For contralateral modulation, a binaural stimulus was used with an unmodulated CF tone of fixed SPL at the ipsilateral ear. For binaural modulation, the value at the interaural time difference (ITD) giving the largest R was taken. Sample sizes, mean values ± SD: AN 316, 0.63 ± 0.12; SBC 11, 0.65 ± 0.11; GBC 47,0.74 ± 0.10; MNTB 24, 0.74 ± 0.08; LSO ipsi 22, 0.88 ± 0.05; LSO contra 10, 0.87 ± 0.11.
, 1995).
; Møller 1976; Palmer 1982). Surprisingly, envelope ITD-sensitivity in LSO becomes unmeasurable at frequencies more than an octave below the highest envelope frequencies found in the AN (Joris 1996). To investigate whether this limitation arises in the LSO afferents, we obtained MTFs for the different types of afferents, examples of which are shown in Fig. 5. For these examples, and in all other afferents examined, the basic shape of the MTF was low-pass. We determined the 3-dB cutoff frequency of each function, which is the modulation frequency at which the gain is 3 dB down from the maximum. The cutoffs of the afferents illustrated (Fig. 5, arrows; for values see legend) are within the range measured in AN fibers of corresponding CF.
View larger version (21K):
[in a new window]
FIG. 5.
Modulation transfer functions for 4 cells in the LSO circuit. Ordinate values are R converted to gain. Positive gain values signify higher modulation depth in the response than in the stimulus. Filled symbols indicate nonsignificant synchronization, arrows indicate 3-dB cutoff point. Insets: averaged spike waveforms with prepotential and the primary-like-with-notch (PLN) histogram used for cell classification. Cells were chosen so that their CFs are approximately equal (within the limits of our sample). Stimulus level (dB) was at SPL giving maximal synchronization at 100 Hz. Cutoff frequency, SPL and CF were: AN: 984 Hz, 4 dB, 19.4 kHz; SBC: 864 Hz, 24 dB, 10.6 kHz; GBC: 1122 Hz, 19 dB, 17 kHz; MNTB: 992 Hz, 29 dB, 18.3 kHz.
View larger version (16K):
[in a new window]
FIG. 6.
Modulation transfer functions to ipsi- and contralateral modulation for 2 LSO cells with the least (left) and most (right) similar cutoff frequencies in the population. Insets: poststimulus time histogram to unmodulated ipsilateral CF tone of 25 ms. Cutoff frequencies for contra- and ipsilateral modulation were 920 and 626 Hz (left) and 541 and 558 Hz (right). SPLs (dB) for ipsilateral modulation were 39 (left) and 14 (right), SPLs for contralateral modulation were; left: 39 (contra)/74 (ipsi), right: 44 (contra)/24 (ipsi). CFs were 20.9 (left) and 18.3 kHz (right).
View larger version (19K):
[in a new window]
FIG. 7.
Cutoff frequencies of modulation transfer functions to monaural modulaton in LSO cells and their afferents. Sample size: AN (138), SBC (5), GBC (35), MNTB (23), LSO ipsi (13), and LSO contra (12).
). A comparison of LSO cutoff frequencies for binaural and monaural modulation is shown in Fig. 8. The monaural values are those of Fig. 7, for the LSO cells that also were studied binaurally. The binaural values are 3-dB cutoff frequencies from synchronization functions to a binaural AM beat stimulus (see Fig. 7B in Joris 1996). The two sets of measures are well correlated (r = 0.82) but virtually all data points lie above the diagonal of equality, i.e., cutoffs to monaural modulation are systematically and significantly higher than the binaural ones (Wilcoxon matched pairs test: P < 104).
View larger version (19K):
[in a new window]
FIG. 8.
Comparison of binaural and monaural 3-dB cutoff frequencies in LSO cells. Monaural values are cutoffs for ipsi- ( 
) and contralateral (
) monaural modulation of Fig. 7, for the subset of cells (n = 13) that also were studied with a binaural AM beat. Binaural values were obtained from synchronization functions to a 1-Hz binaural AM beat stimulus, presented over a range of modulation frequencies. For 2 cells (
and 
) in which phase-locking to the beat became insignificant before a 3-dB value was reached, the highest modulation frequency with significant phase-locking is used as cutoff frequency.
). The functions of LSO cells are more varied and some have more high-pass slope at low modulation frequencies, but overall they are remarkably similar to the afferent responses. Also, there are no indications in our sample of marked differences in shape in the responses to ipsi- and contralateral modulation.
View larger version (19K):
[in a new window]
FIG. 9.
Modulation transfer functions to monaural modulation in LSO cells and their afferents. Ordinate values are normalized to maximum gain; abscissa values are normalized to the 3-dB cutoff frequency at which all functions intersect. To avoid overlap between populations, data of each population are offset by 10 dB, and some functions are clipped at the highest modulation frequencies. For SBCs and LSO cells, all functions reaching a 3-dB cutoff are shown. For GBCs and MNTB cells, a sample of size and CF distribution similar to LSO is shown. Sample size: SBC (5), GBC (14), MNTB (13), LSO ipsi (14), and LSO contra (11). Only data points with statistically significant envelope phase-locking are shown.
View larger version (14K):
[in a new window]
FIG. 11.
A: delays, measured from phase-frequency functions as shown in previous figure. Sample size: AN (153), SBC (7), GBC (40), MNTB (23), LSO ipsi (17), and LSO contra (15). B: ipsi- and contralateral delays, pairwise measured on the same LSO cells, were correlated (r = 0.72). - - -, equality. All data points are corrected for acoustic delay between the driver and a calibration probe close to the eardrum, calculated for each experiment from the phase values of the acoustic calibration (average was 0.4 ms for the AN experiments and 0.28 ms in the CNS recordings, for which different driver assemblies were used).
). The steepness of each function, which provides an estimate of the total delay accumulated between acoustic driver and recording site, was quantified by the slope of a linear regression fitted (all with r2 0.98) to the data points with significant phase-locking. Figure 11A shows the slopes of the linear regressions for the different cell classes as a function of CF. Most striking in these data are the similarity of delays in LSO to ipsi- and contralateral modulation, despite the longer pathway for the contralateral signal including an extra synapse. The mean difference between the ipsi- and contralateral delays (for LSO cells in which both measures were available, n = 15) was 182 µs, indicating that on average, the contralateral inhibition arrived 182 µs later than the ipsilateral excitation. Moreover, the good matching extends to the single cell level because ipsi- and contralateral delays were correlated (r = 0.72), i.e., even though delays differed in magnitude among cells, they tended to be matched within cells (Fig. 11B).
View larger version (28K):
[in a new window]
FIG. 10.
Examples of phase-frequency plots to monaural modulation. These data are the phase portions of the synchronization measurements of the cells shown in Figs. 5 and 6 (right). Data for ipsilateral modulation all converge to a y intercept ~0.25 cycles, which is the stimulus envelope phase measured without time delay (the stimulus envelope started in sine phase). Phase values are uncorrected for acoustic delay. Slopes of linear regressions, corrected for acoustic delay, were: AN, 1.55 ms; SBC, 2.50 ms; GBC, 2.32 ms; MNTB, 3.16 ms; LSO ipsi, 4.75 ms; and LSO contra, 5.32 ms.
). To the extent that differences in delay between populations of the LSO circuit are determined by a fixed conduction delay, the delay values for each population should be DC-shifted relative to its afferents. A least-square fit of the delays in auditory nerve with a power function
= aCFb + d (Anderson et al. 1971; Smolders and Klinke 1986) yielded the parameters (r = 0.93, delay in ms, CF in kHz):
We fit the same function to the other populations of Fig. 11 with d as a free parameter, which provides an estimate of the CF-independent high-frequency asymptote, and obtained 1.8 ms for GBC, 2.2 ms for MNTB, 4.1 ms for LSO responses to contralateral modulation, and 3.9 ms for LSO responses to ipsilateral modulation (1 outlying data point at 519 Hz excluded). (Note that values at 40 kHz have not reached the asymptote but are ~0.5 ms higher).
(1)
or 
) are short and overlap with delays measured in the AN. Delays of PL fibers recorded in the TB (Fig. 12, ) are longer and similar to delays measured on MNTB cells (Fig. 11, 
).
View larger version (14K):
[in a new window]
FIG. 12.
Summary scatterplots for measurements on presumed SBCs. Symbols differentiate cells with primary-like (PL) response pattern, recorded in AVCN (PL-CN) or TB (PL-TB), and AVCN cells with PP. Most (8/11) PL-TB fibers were recorded contralateral to the excitatory ear.
; Joris and Yin 1995). Moreover, the set point around which rate is modulated decreases, so that at high modulation frequencies, firing rate often drops near zero at all ITDs. We made two kinds of observations concerning the origin of this drop in rate. In Fig. 13, we graph the average firing rate counterpart of the modulation transfer functions of Fig. 9, with the addition of a number of presumed SBCs to supplement the small SBC sample. The drop in firing rate at high modulation frequencies also is present in LSO responses to monaural modulation and is strong for contralateral modulation (Fig. 13E) and more varied for ipsilateral stimulation (Fig. 13B). Possible reasons for this drop with increasing modulation frequency would include changes in average firing rate of afferents: a decrease in rate on the excitatory (ipsilateral) side or an increase in rate on the inhibitory (contralateral) side. However, responses from LSO afferents indicate that neither is the case. There is little change in average rate in SBCs (Fig. 13A), consistent with our findings in AN (Joris and Yin 1992). Similarly, rate changes in the contralateral LSO afferents (GBCs in Fig. 13C, MNTB cells in Fig. 13D) are smaller than in LSO and moreover are in the wrong direction. The decrease in rate above ~300 Hz in many MNTB cells should cause decreased inhibition and therefore increased firing rate in LSO cells, contrary to what is observed(Fig. 13E).
View larger version (35K):
[in a new window]
FIG. 13.
Average firing rate for modulation transfer functions shown in Fig. 9. All available data points are shown (up to modulation frequency 
2 kHz), rather than only those with significant synchronization, and different symbols are used for different cells or axons. In addition to the 5 SBCs shown in Fig. 9 (which fullfilled all criteria and are identified here with solid lines and symbols), A includes 9 "presumed SBCs" (see text). Responses to ipsi- (B) and contralateral modulation (E) derived from the same LSO cells are shown with same symbols.
) and may reflect the stronger sensitivity of GBCs and MNTB cells to intensity changes when compared with AN fibersor SBCs.
, 1995 for examples in AN and LSO, respectively). The slope of a linear fit was on average only a few thousandths of a cycle per dB. Expressed in the time dimension (µs/dB), averages were 11.3 (AN, n = 285), 11.4 (SBC, n = 10), 16.7 (GBC, n = 39), 20.3 (MNTB, n = 19), and 27.7 (LSO to ipsilateral modulation, n = 13). In contrast to the negative slope of afferents and LSO-ipsi, the phase of the LSO response to contralateral modulation at 100 Hz increased with SPL: 7.8 µs/dB (n = 6).
View larger version (18K):
[in a new window]
FIG. 14.
A and B: modulation transfer functions at multiple SPLs for a SBC (A, CF = 3.8 kHz) and GBC (B, CF = 28.3 kHz) recorded in CN. Different line types are used to demarcate different functions, and SPL is indicated on each function. C: delays measured from the slope of phase-frequency functions obtained at multiple SPLs in 5 cells. A and B identify delay functions for cells of top panels. Solid symbols indicate cells for which cell class was tentative: the cell with shortest delay values did not show a prepotential but had a PL response (CF: 11.4 kHz), the cell with the longest delay values had a PLN reponse and was recorded in the contralateral TB (CF: 17.2 kHz). CF of the remaining SBC (solid line) was 2.5 kHz. Delay values are corrected for acoustic delay.
) measured in this cell for the two types of carrier (arrows).
View larger version (27K):
[in a new window]
FIG. 16.
ITD-sensitivity in a LSO cell (CF = 6.7 kHz) to a modulated wideband carrier (A) and CF carrier (B). ILD was set to 0 for both types of carrier, and modulation frequency was increased from 50 Hz in 50-Hz steps until modulation became insignificant. Responses are shown over the same range of modulation frequencies (50-450 Hz) and are set apart with alternating line types. Characteristic delay measured (as in Joris 1996 ) over this range was 391 µs for A and 500 µs for B. Responses were obtained for a binaural AM stimulus with an envelope frequency difference (beat) of 1 Hz and are graphed as a function of ITD, calculated from the interaural phase differences in the envelopes of that stimulus (see Joris and Yin 1995). Positive ITD is defined as the contralateral ear leading in time.
), which are of interest in the context of this study because they constitute the other major cell type, besides the bushy cells, of the AVCN and also because they show some structural and physiological resemblances to LSO cells. Both transient and sustained choppers (for definition see legend Fig. 17) had nonmonotonic synchronization-level functions at 100 Hz with higher maximal R values (mean = 0.76 ± 0.071, n = 18) than AN fibers, but their MTFs were less stereotyped with SPL than those of AN fibers and bushy cells and show more of a band-pass characteristic at high SPL, as reported previously (Frisina et al. 1990; Rhode and Greenberg 1994). Cutoff frequencies, measured at the SPL giving the highest gain at 100 Hz, did not show an increase with CF (Fig. 17) and were much lower than in AN or LSO afferents (Fig. 8) particularly at CFs > 10 kHz. The lack of CF dependence suggests a temporal ceiling that is similar for choppers of all CFs.
View larger version (16K):
[in a new window]
FIG. 17.
Modulation transfer function 3-dB cutoff frequencies of choppers (diamonds) compared with AN (+). Cells were classed as choppers on the basis of pattern and regularity of responses to short (25 ms) CF tonebursts. Coefficient of variation (CV: calculated with analysis window of 12-18 ms and bindwidth of 1 ms) was < 0.3 for sustained (Chop S) and 0.5 > CV 0.3 for transient (Chop T) choppers.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Kim et al. 1990; Rhode and Greenberg 1994). For example, CN choppers had high gains but their cutoff frequencies were well below AN values (Fig. 17). In contrast, SBCs and GBCs preserve envelope information over a wide range of modulation frequencies and this range is little affected by SPL (Figs. 7 and 14). These properties, combined with sharp frequency tuning to tones and sustained responsiveness, make these cells ideally suited to supply the binaural system with information regarding amplitude fluctuations over a certain bandwidth of carrier frequencies.
; Rhode and Greenberg 1994; Wang and Sachs 1994) although these studies diverge on the extent of gain increase. Our data indicate the importance of considering CF in such comparisons as populations diverged only clearly below ~7 kHz (Fig. 4). A similar effect can be seen in the data of Wang and Sachs (1994). Comparison of the measurements for AN, PL, PLN, and choppers by Rhode and Greenberg (1994) with our results (legend Fig. 4) for AN, SBCs, GBCs, and choppers shows higher maximal synchronization values in our data (except for GBCs). The average slopes for the synchronization-level functions in the two studies are very similar.
).
; Joris 1996) and are consistent with their good timing properties to pure tone onset and fine-structure (Finlayson and Caspary 1991; Joris and Yin 1995; Tsuchitani 1997). High synchronization values to contralateral stimulation at first may appear more surprising but are qualitatively consistent with a subtraction scheme (see RESULTS). Although other factors, such as postsynaptic temporal smearing of phase-locked inhibitory events (Sanes 1990) or poor synchronization among the MNTB cells converging on the LSO cell, could contribute to a longer inhibitatory postsynaptic potential (IPSP) and consequently narrow period histograms with high synchronization values, our results using ITDs of click stimuli (Joris and Yin 1995) suggest that the effective IPSPs from the contralateral side are short, on the order of 1 ms.
800 Hz (Joris 1996). Such synchronization functions can be viewed as binaural MTFs: responses only can be modulated by envelope interaural phase differences if both ipsi- and contralateral input pathways carry envelope phase information. We therefore compared the phase-locking range in three steps of the LSO circuit: in LSO afferents, in LSO cells to monaural modulation, and finally in LSO cells to binaural modulation. The phase-locking ranges of LSO afferents are essentiallyAN-like (Fig. 7); the limited range for ITD-sensitivity in LSO therefore must derive from postsynaptic limitations such as poor synchronization across converging MNTB or SBC afferents, dendritic filtering, or temporal summation of subthreshold events. Monaural MTFs measured on LSO cells did show a more limited range of phase-locking than LSO afferents (Fig. 7), but, surprisingly, not as limited as the range for ITD-sensitivity (Fig. 8). This discrepancy may originate partly from our experimental and analytic procedures. R values to a binaural beat stimulus are lower (<0.6, see Fig. 7A in Joris 1996) and noisier than monaural R values at the modulation frequency (>0.6, Fig. 4), which tends to reduce the cutoff value (see also Batra et al. 1997b). Also, the decrease in average rate with modulation frequency was stronger with binaural than with monaural modulation, probably reflecting the smaller ILD range used binaurally (±20 dB) than monaurally (30 to 35 dB for contralateral modulation and usually no contralateral stimulus during ipsilateral modulation). Nevertheless, there are clear examples of cells that phase-lock monaurally to modulation frequencies at which there is no detectable ITD-sensitivity. Thus not all LSO cells transform the full-frequency range of temporal information supplied by its afferents into ITD-sensitivity.
; Wu and Kelly 1991) and 2) distal (SBC) versus proximal (MNTB) placement of the synaptic terminals (Cant 1984; P. H. Smith, P. X. Joris, and T.C.T. Yin, unpublished data). The observation that cutoff frequencies vary widely in different cells, but within a cell, tend to be similar for ipsi- and contralateral modulation and are correlated with the cutoff for binaural modulation (Fig. 8) suggest that the frequency limitation is at a postsynaptic stage that influences both monaural channels to the same degree.
), a sustained, unmodulated MNTB input inhibits the LSO cell more effectively than a modulated MNTB input. The rate decrease for ipsilateral modulation is harder to explain but also more variable. A possible explanation is complementary to the one just offered for contralateral modulation. MNTB cells provide a base level of inhibitory input (their average spontaneous rate was 29 spikes/s) to LSO. This input likely blocks modulated excitatory input (at low ipsilateral modulation frequencies) less effectively than sustained excitatory input (at high modulation frequencies).
). We found that this similarity extends to responses to envelope modulation in terms of gain, delay, shape, and cutoff frequency of modulation transfer functions. The response parameter that has attracted most attention in interaural comparisons is latency. Previous studies with tones and clicks (Boudreau and Tsuchitani 1968; Caird and Klinke 1983; Sanes and Rubel 1988; Tsuchitani 1988, 1997) suggested similar latencies for inhibition and excitation because even the first spike in response to ipsilateral stimulation can be inhibited by a simultaneously gated stimulus to the contralateral ear. However, direct measurement of latency for the inhibitory ear is difficult with such stimuli. We estimated delays by measuring the slope of the phase-frequency relationship in each cell (Figs. 10 and 11). Such delays do not show a decrease with increasing SPL (Fig. 14C), unlike traditional measures of onset latency, but consistent with the notion that they are determined mainly by fixed conduction delays. Similar observations using a different technique were made by Møller (1975).
). The mean characteristic delay for all cells was 200 µs, which is identical to the difference in asymptotic delay measured here (Fig. 18: 4.1 and 3.9 ms for contra- and ipsilateral modulation, respectively) and close to the mean difference in contra- and ipsilateral delays (182 µs). Thus on average, the contralateral signal is delayed slightly, by 0.2 ms, relative to the ipsilateral signal. Even more striking is the observation that the delays are correlated (Fig. 11B), suggesting that the matching extends even to the single cell level. Small characteristic delays, but with an opposite bias, also were found by Batra et al. (1997a) in cells with ITD-sensitivity similar to LSO cells.
View larger version (17K):
[in a new window]
FIG. 18.
Summary diagram of delay measurements (in ms) on a tracing of a coronal section through the superior olivary complex. Values are the asymptotic or CF-independent part of a power function fit to the data of Fig. 11A (see text). Italic numbers on bottom give differences in delay between cell classes. Long delay between AN and LSO to ipsilateral modulation (2.9 ms) probably derives from the small diameter and circuitous path of SBC axons, and their termination on distal LSO dendrites. Comparatively short delay between AN and LSO to contralateral modulation (3.1 ms in total) derives from fast axonal conduction and axosomatic placement of terminals along each relay point.
: 8-15 µm; Spangler et al. 1985: 8-15 µm; Spirou et al. 1990; Smith et al. 1991; van Noort 1969: 8-12 µm), smaller for MNTB (Spangler et al. 1985: 5-6 µm; P. H. Smith, P. X. Joris, and T.C.T. Yin, unpublished results: 4-6 µm), and apparently yet smaller for SBCs (Brownell 1975: 3-5 µm; Smith et al. 1993: 2.5-5 µm for low-CF SBC axons; Warr 1966: 1.8 µm; van Noort 1969: 4 µm) though we are not aware of quantitative measurements on identified high-CF SBC axons projecting to LSO. Another factor in the long ipsilateral delay may be path length. Labeled low-CF SBCs project to LSO via a variable and circuitous route (Smith et al. 1993). Moreover, the ipsilateral signals likely incur a substantial dendritic delay because of the distal placement of the SBC terminals on LSO cells (Cant 1984).
) that average rate of LSO cells is affected by both ITDs and ILDs. It is well known that changes in SPL affect response onset latency, so that ILDs may affect firing rate not only by changes in firing rate in afferent pathways but also by changing time of arrival of afferent spikes at the site of convergence. This is the basis for the latency hypothesis (Jeffress 1948). Despite the absence of change with SPL in the slope of the phase-frequency functions, the response phase at a single modulation frequency decreases linearly with increasing SPL, as we previously described in AN and LSO. This is likely due to adaptation, which shifts spikes to earlier phase values of the modulation cycle. The change in response phase was small (~10-30 µs/dB) but possibly significant given the physiological ranges of ITDs (±400 µs for cat) (Roth et al. 1980) and ILDs (±30 dB) (Irvine 1987; Musicant et al. 1990). However, the functional importance of the latency-SPL interaction for AM responses in LSO is questionable because of the large "direct" effect of SPL and ILD on firing rate (Joris and Yin 1995). The interaction may be more important for transient responses (Joris and Yin 1995; Pollack 1988; Yin et al. 1985; Irvine et al. 1998).
) and would be categorized physiologically as presumed SBC if recorded in the TB, or SBC if recorded from CN [where GBCs also can display large PPs (Bourk 1976)]. Another possible bias between SBC and GBC populations may have been introduced by the somatic (SBC) versus axonal (GBC) recordings. For example, precision of phase-locking to pure tones was found to be significantly better in axonal recordings from bushy cells when compared with most values reported for extracellular recordings in CN (Joris et al. 1994a). However, the similarity in findings between MNTB cells and GBCs, as well as between the SBC and presumed SBC populations, argue against the significance of any such effects on envelope synchronization.
), but unphysiologically large ITDs are necessary to move the intracranial image away from midline (Bernstein and Trahiotis 1985). Physiologically, rate is much more dependent on ILD than on ITD if stimuli are restricted to the physiological range (Joris and Yin 1995). The weakness of the ITD-sensitivity is in stark contrast to the dramatic dimensions of the functional and morphological features, e.g., the endings of globular bushy cells on MNTB cells, the calyces of Held, are probably the largest terminals in the mammalian CNS (Jean-Baptiste and Morest 1975). On the other hand, these features seem unnecessary or even disadvantageous for ILD-sensitivity. It can be argued that an ILD-extracting circuit should optimally receive inputs that integrate over time and perhaps frequency because ILDs are a complex function of frequency at each spatial position (Musicant et al. 1990; Rice et al. 1992). Other cochlear nucleus cell types seem better suited to convey information on SPL needed to process ILDs (Rhode and Smith 1986; Shofner and Dye 1989). If not to enable extraction of ITDs or static ILDs, what then is the benefit of these specializations? We suggest that specializations in the LSO circuit optimize preservation of envelope information and comparison of instantaneous amplitude fluctuations at the two ears.
; Jeffress 1948; Yin and Chan 1990). A similar view for LSO would hold that the specializations to match the input pathways in terms of timing and spectral integration, as well as the presence of IE-interaction, have evolved to allow calculation of ITDs and that ILD-sensitivity is a byproduct of this evolution. Given that physiologically and psychophysically at high frequencies ILDs are a much more potent cue than ITDs, this view is untenable.
) and of differences in monaural delays (Fig. 11) relative to the long delays required to carry the signals to the LSO argues that the LSO circuit embodies the latter solution. Through its specializations, this system is able to extract ILDs, and its output still contains information about temporal modulations in amplitude and frequency.
; DeAngelis et al. 1991). Such detectors can track disparity of a time-varying (e.g., moving) stimulus only if the two eyes supply temporal information matched for the two eyes. Being a slower sensory system, the timing requirements in vision are less demanding than in audition, so that large temporal mismatches are tolerated before stereoscopic perception is lost (50 ms) (review in Howard and Rogers 1995). An example of the effect of interocular timing is found in the Pulfrich illusion, in which motion in the fronto-parallel plane appears to be in depth if one eye is covered with a light-attenuating filter. The filter introduces a neural temporal delay, which shifts the interocular phase at which binocular cells discharge maximally (Carney et al. 1989). We surmise that this time sensitivity should be viewed as a necessary outcome of a spatial disparity detector with a time resolution scaled to the demands of visual processing. Similarly, we suggest that the good match of interaural delays in the LSO has evolved to enable the extraction of ILDs at corresponding points in frequency and time of the acoustic waveform in both ears. In a sense, there is a time correspondence problem for ILD computation, like there is a spatial correspondence problem for stereoscopic vision. If the monaural signals would have been delivered by sluggish channels preserving little timing information, the requirement for temporal matching of ipsi- and contralateral signals would have been less stringent, as for binocular cells. However, the afferents of the LSO contain envelope-related timing information over the full range available in the AN. Thus it is important that the interaural timing in this circuit is controlled precisely.
by virtue of the sigmoidal ILD function and activity summed across cellsreinforcing the need for a multiplexing strategy.
). Also, it is plausible that preservation of modulation information after binaural interaction is needed to allow integration of cues at a later stage. For example, envelopes correlated across frequency regions improve the lateralization of high-frequency complex sounds (Saberi 1995). Another example is the improved speech intelligibility in a noisy environment by integration of modulation and binaural cues, as exemplified by a computer algorithm (Kollmeier and Koch 1994). Third, this series of papers has emphasized responses to ongoing ITDs in AM stimuli, but accuracy in localizing high-frequency transients ("snapping twigs") or detection of onset time differences also may require the specializations for time found in the LSO circuit (Caird and Klinke 1983; Joris and Yin 1995; Irvine et al. 1998). Finally, it has been suggested that binaural cues may be coded by the timing patterns of LSO discharges rather than by rate of firing (Tsuchitani 1997).
) that our interpretation does not preclude the usefulness of high-frequency ITDs for localization of sounds. The core of our argument is that this capability is too weak and occurs under too stringent conditions to warrant the powerful morphological and physiological specializations that are found in this circuit. Whether computational demands along the lines suggested necessitate these specializations remains to be demonstrated behaviorally and physiologically.
| |
ACKNOWLEDGEMENTS |
|---|
The authors thank R. Batra, S. Kuwada, A. R. Palmer, and P. H. Smith for comments on the manuscript. We gratefully acknowledge the assistance of I. Siggelkow, J. Meister, J. A. Ekleberry (for histology), R. Kochhar, J. Sekulski (for software), and G. Meulemans (for photography).
This work was supported by National Institute of Deafness and Other Communications Disorders Grant DC-00116.
| |
FOOTNOTES |
|---|
Address for reprint requests: P. X. Joris, Division of Neurophysiology, Medical School, University of Leuven, Campus Gasthuisberg, B-3000 Leuven, Belgium.
Received 7 May 1997; accepted in final form 5 August 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Mc Laughlin, M. van der Heijden, and P. X. Joris How Secure Is In Vivo Synaptic Transmission at the Calyx of Held? J. Neurosci., October 8, 2008; 28(41): 10206 - 10219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dreyer and B. Delgutte Phase Locking of Auditory-Nerve Fibers to the Envelopes of High-Frequency Sounds: Implications for Sound Localization J Neurophysiol, November 1, 2006; 96(5): 2327 - 2341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Siveke, M. Pecka, A. H. Seidl, S. Baudoux, and B. Grothe Binaural Response Properties of Low-Frequency Neurons in the Gerbil Dorsal Nucleus of the Lateral Lemniscus J Neurophysiol, September 1, 2006; 96(3): 1425 - 1440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Tollin and T. C. T. Yin Interaural Phase and Level Difference Sensitivity in Low-Frequency Neurons in the Lateral Superior Olive J. Neurosci., November 16, 2005; 25(46): 10648 - 10657. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Stecker Rate-Limited, But Accurate, Central Processing of Interaural Time Differences in Modulated High-Frequency Sounds. Focus on: "Neural Sensitivity to Interaural Envelope Delays in the Inferior Colliculus of the Guinea Pig" J Neurophysiol, June 1, 2005; 93(6): 3048 - 3049. [Full Text] [PDF]
|
|
|
|
|
|
S. J. Griffin, L. R. Bernstein, N. J. Ingham, and D. McAlpine Neural Sensitivity to Interaural Envelope Delays in the Inferior Colliculus of the Guinea Pig J Neurophysiol, June 1, 2005; 93(6): 3463 - 3478. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. H. G. Louage, M. van der Heijden, and P. X. Joris Enhanced Temporal Response Properties of Anteroventral Cochlear Nucleus Neurons to Broadband Noise J. Neurosci., February 9, 2005; 25(6): 1560 - 1570. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. H. G. Louage, M. van der Heijden, and P. X. Joris Temporal Properties of Responses to Broadband Noise in the Auditory Nerve J Neurophysiol, May 1, 2004; 91(5): 2051 - 2065. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. X. JORIS, C. E. SCHREINER, and A. REES Neural Processing of Amplitude-Modulated Sounds Physiol Rev, April 1, 2004; 84(2): 541 - 577. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Sterbing, W. R. D'Angelo, E.-M. Ostapoff, and S. Kuwada Effects of Amplitude Modulation on the Coding of Interaural Time Differences of Low-Frequency Sounds in the Inferior Colliculus. I. Response Properties J Neurophysiol, November 1, 2003; 90(5): 2818 - 2826. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Kopp-Scheinpflug, K. Fuchs, W. R. Lippe, B. L. Tempel, and R. Rubsamen Decreased Temporal Precision of Auditory Signaling in Kcna1-Null Mice: An Electrophysiological Study In Vivo J. Neurosci., October 8, 2003; 23(27): 9199 - 9207. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. X. Joris Interaural Time Sensitivity Dominated by Cochlea-Induced Envelope Patterns J. Neurosci., July 16, 2003; 23(15): 6345 - 6350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Tollin The Lateral Superior Olive: A Functional Role in Sound Source Localization Neuroscientist, April 1, 2003; 9(2): 127 - 143. [Abstract] [PDF]
|
|
|
|
 |
|
 H. M Brew, J. L Hallows, and B. L Tempel Hyperexcitability and reduced low threshold potassium currents in auditory neurons of mice lacking the channel subunit Kv1.1 J. Physiol., April 1, 2003; 548(1): 1 - 20. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Macica, C. A. A. von Hehn, L.-Y. Wang, C.-S. Ho, S. Yokoyama, R. H. Joho, and L. K. Kaczmarek Modulation of the Kv3.1b Potassium Channel Isoform Adjusts the Fidelity of the Firing Pattern of Auditory Neurons J. Neurosci., February 15, 2003; 23(4): 1133 - 1141. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D.R.F. Irvine, V. N. Park, and L. McCormick Mechanisms Underlying the Sensitivity of Neurons in the Lateral Superior Olive to Interaural Intensity Differences J Neurophysiol, December 1, 2001; 86(6): 2647 - 2666. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Snyder, M. Vollmer, C. M. Moore, S. J. Rebscher, P. A. Leake, and R. E. Beitel Responses of Inferior Colliculus Neurons to Amplitude-Modulated Intracochlear Electrical Pulses in Deaf Cats J Neurophysiol, July 1, 2000; 84(1): 166 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. S. Krishna and M. N. Semple Auditory Temporal Processing: Responses to Sinusoidally Amplitude-Modulated Tones in the Inferior Colliculus J Neurophysiol, July 1, 2000; 84(1): 255 - 273. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Gardner, L. O. Trussell, and D. Oertel Time Course and Permeation of Synaptic AMPA Receptors in Cochlear Nuclear Neurons Correlate with Input J. Neurosci., October 15, 1999; 19(20): 8721 - 8729. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kuwada and R. Batra Coding of Sound Envelopes by Inhibitory Rebound in Neurons of the Superior Olivary Complex in the Unanesthetized Rabbit J. Neurosci., March 15, 1999; 19(6): 2273 - 2287. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. X. Joris and P. H. Smith Temporal and Binaural Properties in Dorsal Cochlear Nucleus and Its Output Tract J. Neurosci., December 1, 1998; 18(23): 10157 - 10170. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Burger and G. D. Pollak Analysis of the Role of Inhibition in Shaping Responses to Sinusoidally Amplitude-Modulated Signals in the Inferior Colliculus J Neurophysiol, October 1, 1998; 80(4): 1686 - 1701. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
