| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Coleman Memorial Laboratory, W. M. Keck Center for Integrative Neuroscience, Department of Otolaryngology, University of California, San Francisco, California
Submitted 7 May 2007; accepted in final form 6 September 2007
 |
ABSTRACT |
|---|
|
|
|
INTRODUCTION |
|---|
|
; Inouye 2000; Middlebrooks et al. 2002). In vision, the identity of an object is often invariant to spatial translation and retinal location. This may explain why receptive field (RF) features, such as orientation preference or ocular dominance, are expressed across large spans of the retinotopic extent of the primary visual cortex (Bosking et al. 2002; Hubel and Wiesel 1977; Hübener et al. 1997; Issa et al. 2000; Swindale et al. 2000). By contrast, the auditory system processes behaviorally relevant events that are often linked to specific frequency regions and thus segments along the cochlear receptor surface (Reser et al. 2000; Suga 1988). Many RF properties of primary auditory cortex (AI) neurons such as binaural interaction and intensity tuning form spatial clusters embedded in the global tonotopic organization (Eggermont 2001; Imig and Adrián 1977; Nelken 2002; Phillips et al. 1994). Some of them, such as binaural interaction classes, are frequency dependent, and their distributions can differ along the cortical tonotopic axis (Kelly and Sally 1988; Razak and Fuzessery 2002; Reser et al. 2000). Further segregation of functional AI properties exists for frequency selectivity or spectral integration (Cheung et al. 2001; Merzenich et al. 1975; Read et al. 2001; Recanzone et al. 1999; Schreiner and Mendelson 1990; Schreiner et al. 2000). Narrow-band regions contain neurons that are sharply tuned to frequency, and neighboring neurons have similar frequency preference. Broad-band regions contain broadly tuned neurons, and/or nearby cells have significant scatter in their preferred frequencies (Schreiner and Sutter 1992; Schreiner et al. 2000). The aim of this study was to determine whether frequency decomposition, expressed in the cortical tonotopic gradient, systematically interacts with spectral integration. We found that the local variation in spectral integration capacity in cat AI is frequency specific and that distinct spectral bandwidth modules are most strongly expressed in neurons with characteristic frequencies (CFs) between 5 and 20 kHz. Moreover, the magnitude of the cortical frequency gradient covaries with the local spectral bandwidth distribution. Together, these observations establish interactions and local constraints on the cortical expression of two basic auditory processing parameters. |
|
METHODS |
|---|
|
Experiments were conducted on four right hemispheres from one male and three adult female cats. All protocols were approved by the University of California San Francisco Committee on Animal Research in accordance with federal guidelines for care and use of animals in research. Animals were sedated by intramuscular injections of a mixture of ketamine (22 mg/kg) and acepromazine (0.11 mg/kg). After venous cannulation, sodium pentobarbital (15–30 mg/kg) was administered and supplemented as required throughout the surgical procedure. After tracheotomy, a craniotomy exposed the ectosylvian gyrus. The dura mater was partially removed, and the cortical surface was covered with thick silicone oil. Before commencing the electrophysiological recordings, sodium pentobarbital anesthesia was replaced with a continuous intravenous infusion of a mixture of ketamine (2–10 mg/kg/h) and diazepam (0.05–0.2 ml/kg/h) in lactated Ringer (1–3 ml/kg/h). To prevent edema and mucus secretion, dexamethasone (1.2 mg/kg sc) and atropine sulfate (0.04 mg/kg sc) were injected at regular intervals. Body temperature was monitored and maintained by a water heating pad at 37 ± 1°C. Electrocardiogram and respiration were monitored continuously during the surgery and recording procedures. Because recordings lasted for 3–4 days, cephalosporin (11 mg/kg iv) was administrated to prevent wound infection.
Acoustic stimulus
Experiments were conducted in a double-walled, anechoic chamber (Industrial Acoustic, Bronx, NY). Pseudorandomized tone bursts (675) at different frequencies (45 different frequencies in 3–6 octaves) and sound level (70-dB range in 5-dB steps) were presented to the left ear sealed by a STAX speaker. The system frequency transfer function was nearly flat (±6 dB) for frequencies 
14 kHz and attenuated 10 dB/octave >14 kHz. As a consequence, CFs >
28 kHz tend to be underestimated (see following text). Sound stimuli of 50-ms duration including 3-ms linear rise and fall time were generated at an interval of 400–750 ms by a microprocessor (TMS32010, 16 bit resolution and 120 kHz D/A sampling rate). Pure tone or white noise bursts were used as search stimuli.
Brain mapping
A video picture of the auditory cortex (AC) surface was captured and digitized with a CCD digital camera (Cohu, San Diego, CA). A parylene-coated tungsten microelectrode (0.5–1.5 M
, Micro Probe, Gaithersburg, MD) was advanced perpendicular to the AC surface with a hydraulic microdrive (David Kopf Instruments, Tujunga, CA). Recordings of single- and multiunit activity were performed at depths of 750–1,050 µm (corresponding to layers IIIb and IV) (Huang and Winer 2000). Each penetration was marked on the digitized picture using Canvas software (Deneva, Miami, FL). The marked sites were used to reconstruct tessellation maps of the recording area.
The anterior border of AI was determined by CF-gradient reversal for all four cases (Imaizumi et al. 2004; Knight 1977). The dorsal and ventral extent of AI was estimated by the occurrence of multipeaked tuning curves (Sutter and Schreiner 1991), long onset latencies (He and Hashikawa 1998), increased response thresholds, and loss of strict tonotopy (Schreiner and Cynader 1984). The posterior AI border could not be determined because it is inside the posterior ectosylvian sulcus (Reale and Imig 1980). The dorsal and ventral borders determined by physiological criteria were well matched to anatomical borders determined by immunostaining with SMI-32, an antibody that recognizes neurofilaments of apical dendrites of AC pyramidal neurons in fixed coronal sections (two cases: AMC6 and ADX4; data not shown).
Data analysis
Data were analyzed using the MATLAB (MathWorks, Natick, MA) platform. StatView (SAS Institute, Cary, NC) was used for statistical analysis. CF (the frequency at which a neuron or a neuron cluster produces sound-evoked spikes at lowest threshold level), response threshold at CF, and spectral bandwidth at 10 and 40 dB above threshold (Q10 and Q40: CF divided by bandwidth; the higher the Q value, the narrower the spectral bandwidth) were determined from 161 to 206 AI sites per case (Table 1). For determining the spatial Q40 modulation index (Q40 MI; Fig. 4), Q40 values were three-point-averaged along the dorso-ventral dimension within 4-kHz-wide frequency bins. Q40 MI is defined as the difference between the largest and smallest Q40 values (average of 2–4 values) along an isofrequency contour. The local frequency gradient (the first spatial derivative of the CF map) was first determined from an interpolated true-value representation of CFs by Surfer 8.0 (Golden Software, Golden, CO) and then constrained to the actual recording locations.
|
|
|
|
|
|
|
RESULTS |
|---|
|
) by single- and multiunit extracellular recording techniques. Frequency decomposition of auditory stimuli and spectral integration are two fundamental features of auditory processing and contribute to various auditory behaviors and perception (Schreiner et al. 2000; Suga 1988). To examine the relationship between these two properties of spectral analysis, we mapped the spatial distribution of local spectral selectivity measures, Q10 and Q40, over as broad a frequency range as experimentally feasible. Q10 and Q40 values reflect spectral bandwidth (or excitatory RF size) near and away from response threshold, respectively, and, thus reveal different contributions to spectral integration related to thalamocortical input and local cortical transformations (Suga 1995; Sutter et al. 1999). Spectral integration analysis
Cochleotopic representation is a conspicuous physiological feature in AI (Merzenich et al. 1975), and is expressed as a smooth gradient of the CF. Tessellation maps (Fig. 1, A and B) capture an undistorted view of the spatial organization of AI with a postero-anterior tonotopic gradient from low to high frequencies (red to purple colors). The spatial pattern of the Q10 and Q40 values (Fig. 1, C–F) reveals a nonuniform distribution of spectral integration along the frequency gradient as well as within the isofrequency domain. Blue and red polygons correspond, respectively, to recording sites with broad and narrow spectral bandwidths. The spatial distribution of the Q10 values differs from that of the Q40 values. Large high-Q10 clusters appear in the mid to high-frequency range and rarely in the lower-frequency region (Fig. 1, C and D). By contrast, high Q40 clusters are spatially more confined and are limited to the mid-frequency region (Fig. 1, E and F). There, high Q40 clusters are flanked dorsally and ventrally by low Q40 clusters, resulting in a systematic modular organization of spectral bandwidth along these isofrequency contours (Read et al. 2001; Schreiner et al. 2000). In the lower- and higher-frequency regions, however, Q40 values appear more homogeneously distributed (Fig. 1, E and F).
Our frequency mapping spanned more than five octaves in three cases and four in the other (Table 1). Q values pooled from the four cases showed a similar distribution as individual tessellation maps (Fig. 2, A and B). Q10 and Q40 values <5 kHz are low and are homogeneously distributed, whereas Q values in the mid-frequency range have a wide distribution including low and high values. Q40 values in the high-frequency range are more confined and, with some exceptions, have more homogeneous distribution than in the mid-frequency range (Fig. 2B; note difference in scales). The mean Q10 and Q40 values in 1/2-octave bands as a function of CF reveal the global frequency dependence of AC spectral integration (Fig. 2, C and D). Statistically, the highest mean Q values lay between 9 and 33 kHz for Q10 and between 9 and 21 kHz for Q40 (central white region in Fig. 2, C and D). Significantly lower Q10 values were found <3.8 kHz (F test adjusted by the sequential Bonferroni correction). Gray background in Fig. 2, C and D, indicates transition ranges, i.e., these CF ranges showed no statistically significant differences to either low or high Q regions (white areas).
The range of Q values in different frequency bands is not uniform. Figure 3, A and B illustrates the distribution of Q40 values from individual cases (
). Shown are fitted models of the CF dependence by a second-order nonparametric, local regression (—) to illustrate the mean Q40 distribution. The individual cases indicate a frequency dependence of mean and variance in the Q40 distribution. Because a modular organization of spectral bandwidths appears most prominently expressed for Q40 (Fig. 1), we further quantified this distribution in two ways. First, we determined the range of Q40 values as a function of CF. This was computed as Q40 residuals by a second-order nonparametric, local regression fit to the Q40 distributions to remove the frequency-dependence of the mean Q40 values (Fig. 3, A and B, —). The SD of Q40 residuals was quantified in contiguous 4-kHz bands across the mapped frequency range (Fig. 3, C and D). For CF ranges between 5 and 21 kHz, the SDs of Q40 residuals were significantly higher than for the 1- to 5- and 25- to 29-kHz ranges (F test adjusted by the sequential Bonferroni correction). This indicates a reduction in the local variance of Q40 values in the low and high CF ranges relative to the mid-CF range.
The spatial bandwidth organization was further quantified by assessing bandwidth modularity strength within isofrequency contours. Spatial modulation of Q40 values along the dorso-ventral (isofrequency) dimension was quantified as Q40 MI (see METHODS) in contiguous 4-kHz bands across the mapped frequency range (Fig. 4A) . Along many isofrequency contours, two discrete maxima of Q40 values correspond to two narrow-bandwidth modules embedded in regions of broader bandwidth (Fig. 1E for case AWD3). Figure 4B depicts the population of Q40 MI values. ANOVA reveals a significant difference (P = 0.002, F = 5.4) in the expression of bandwidth modularity between different frequency regions. Specifically, CFs between 5 and 21 kHz show a significantly larger strength of spatial organization than the 1- to 5-and 25- to 29-kHz ranges (P < 0.01: Fisher's protected least significant difference test). This suggests that spectral bandwidth modularity is heterogeneously expressed across different CF ranges. Between CFs of 5 and 21 kHz, Q40 values are spatially organized along isofrequency contours (Fig. 4B). By contrast, no consistent spatial modulation of Q40 values is discernable <5 and >20 kHz along the dorso-ventral axis. As a consequence of the wider range of Q40 values and the spatial Q40 organization along mid-frequency isofrequency contours, the local Q gradient is substantially shallower in the low- and high-frequency regions.
The majority of these recordings were made in the mid-frequency regions due to the geometry of sulci and large exposure of this region on the cortical surface (Fig. 1). Figure 5A illustrates the distribution of sampling in the different CF ranges for the whole population. About 63% of the recordings originated in the mid-frequency range (5–20 kHz), whereas 11 and 25% of the recordings were made in either the low- or high-frequency range, respectively. Given the prevalence of high Q40 values in the mid-frequency range, the sampling difference between the three frequency regions may create a bias in the statistical evaluation. Monte Carlo analysis was performed to assess whether estimates of the mean Q40 values were influenced by these sampling differences. Eighty Q40 values corresponding to the number of recorded low-frequency units (Fig. 5A) were randomly chosen 10,000 times from the mid-frequency range dataset (463 units). The resulting distribution of mean Q40 values was compared with the actual mean values obtained for the low- and high-frequency regions (Fig. 5B). As shown in Fig. 5B, the mean Q40 values for the low- and high-frequency ranges (
) were significantly (P < 0.0001 for low-frequency range and P = 0.0011 for high-frequency range) below the distribution of those for the randomized mid-frequency samples. Therefore higher mean Q40 values in the mid-frequency range are not influenced by oversampling. Rather they accurately reflect the nature of sound frequency integration for that particular frequency range.
|
Inspection of Fig. 1, A and B, suggests that the tonotopic gradient across AI is not uniform. Replotting the CF map as an interpolated but true-value representation (Fig. 6A) with overlaying isofrequency contours (1/3-octave intervals) confirms that the interval between neighboring isofrequency contours is not uniform across AI. In other words, the CF gradient is variable. Overlaying the isofrequency contours onto the Q40 distribution (Fig. 6B) suggests a wider frequency contour spacing in high Q40 regions (red regions) than in broadly tuned (blue) regions. Plotting the local slope of the tonotopic map (in octaves/mm; the inverse of the cortical frequency magnification factor) reveals distinct regions of high and low gradient magnitude, reflecting the local variations in isofrequency contours spacing (Figs. 6C and 7, A, C, and E). Overlaying the Q40 distribution onto the CF gradient map illustrates that high Q40 values are concentrated in the mid-frequency region with a shallow CF gradient (Figs. 6D and 7, B, D, and F) and the high Q40 regions are surrounded by patches of particularly steep frequency gradients (with the exception of the high-frequency side; see following text).
|
|
) reveals that the cortical gradient is substantially steeper <5 kHz, indicating a smaller than expected cortical magnification factor. The CF gradient between 9 and 20 kHz matches the cochlear gradient adjusted to the extent of AI. For high CFs, the gradient is the shallowest. However, estimation of CFs >28 kHz was influenced by the sensitivity roll-off in the loudspeaker transfer function (see METHODS) possibly leading to underestimation of the actual CF and, thus underestimating the slope of the CF gradient (super-threshold Q values, especially Q40, are less affected by the roll-off). The transition between steep and shallow gradient ranges (Fig. 8B, gray area: no statistical difference with either high- or low-CF gradient regions) is centered around 7 kHz. This corresponds well to the transition ranges between high and low Q10 (midpoint of transition range: 6 kHz) and Q40 (6.5 kHz) values (see Fig. 2, C and D).
|
|
|
|
DISCUSSION |
|---|
|
Many response features in primary sensory cortex have an orderly periodic, uniform distribution (Bosking et al. 2002; Hubel and Wiesel 1977; Hübener et al. 1997; Issa et al. 2000; Swindale et al. 2000). Examples include orientation selectivity, spatial frequency, and ocular dominance in the primary visual cortex. We find that this principle does not extend to the spectral bandwidth organization in cat AI. Previously, spectral bandwidth organization in cat AI had only been assessed in a part of the mid-frequency range (Read et al. 2001; Schreiner and Mendelson 1990). The current findings suggest that cat AI comprises three frequency regions (<5, 5–20, and >20 kHz) distinguished by the range of Q40 values and the expression of spatial organization of spectral bandwidths.
The heterogeneity of spectral integration properties across AI is in contrast to psychophysically determined spectral integration that is relatively constant at a "critical bandwidth" of 1/3 octave throughout the cat hearing range (Ehret and Schreiner 1997; Nienhuys and Clark 1979; Pickles 1975). Differences in the spatial distribution between Q10 and Q40 values suggest differences in the neural mechanisms underlying the neural cluster responses that are related to peripheral representation and thalamocortical projections as well as to the RF construction in AC (Miller et al. 2001; Suga 1995; Sutter et al. 1999). Spectral bandwidth is already influenced by cochlear tuning properties (Liberman 1978; Narayan et al. 1998) and is reflected in subsequent processing stations. However, the spectral integration differences described here for three frequency regions likely include higher-order processing principles, presumably related to specific behavioral tasks (e.g., Razak and Fuzessery 2006; Suga 1988) and reflecting neuroanatomical connectivity principles (Prieto et al. 1994; Read et al. 2001).
Nonuniform distributions of spectral integration properties may be seen in other species such as the ferret (Shamma et al. 1993), owl monkey (Recanzone et al. 1999), squirrel monkey (Cheung et al. 2001), and marmoset (Philibert et al. 2005). However, the expression of spectral integration properties in these species varied from case to case and revealed no clear frequency dependence. A modular organization of spectral bandwidth in cat anterior auditory field (AAF), the other primary field, is highly idiosyncratic and frequency-independent as well (Imaizumi et al. 2004). Functional segmentation of the tonotopic domain is also found in the moustached bat AI: a primary area for object velocity and angle detection (DSCF area) occupies a narrow frequency range (CF = 61–63 kHz) devoted to processing of a dominant frequency in the bat's biosonar signal (Suga 1988). Thus three different frequency regions in the moustached bat AI are related to specific auditory tasks. By analogy, the three distinct frequency regions in cat AI may reflect specific functional tasks. An anatomical study in guinea pig AI also suggested functional distinctions between low- and high-frequency regions based on differences in their projections (Wallace et al. 2002). A prospective candidate task for the mid-frequency range may be the processing of spectral sound localization cues. Monaural cues to locate sound sources in the form of spectral notches are created by filtering properties of the head and pinnae (Yin and May 2005; Young and Davis 2002). The first spectral notch in the cat head-related transfer functions occurs predominantly at frequencies between 5 and 20 kHz depending on the sound source direction (Yin and May 2005; Young and Davis 2002). The frequency range closely matches that for the occurrence of high and low Q40 modules. However, unlike the dorsal cochlear nucleus (Young and Davis 2002), no clear evidence yet links cortical function to notch processing. Binaural summation and suppression bands also appear to be more strongly expressed in the mid- to high-frequency ranges, whereas low-frequency regions are dominated by summation interactions (Imig and Adrián 1977; Imig and Brugge 1978), suggesting the potential link between spectral shape processing and binaural interactions. However, preliminary data from an AI mapping study revealed no systematic relationship between spectral bandwidth and binaural interaction (Teng et al. 2006). A recent study in cat AI has shown that prolonged sound exposure of juvenile cats to the mid-frequency range of 5- to 20-kHz shapes the functional organization (Noreña et al. 2006), thus indicating a particular susceptibility of this region for environmental and behavioral influences.
The functional interpretation of these findings is confounded by systematic variations of other RF parameters with frequency and within the isofrequency domain. In this scenario, the combinations of different RF properties present in each neuron, such as spectral bandwidth, response threshold, monotonicity, etc., will affect the response strength and disclose important principles of population activity and consequences of multidimensional influences on stimulus representation. The current study focuses on simple RF properties that, in themselves, do not provide a very satisfactory answer to the question of how a given stimulus is represented. However, it allows a first assessment whether regional functional differentiations exist and how that topography may be related to other functional topographies and spatial patterns of thalamocortical projections.
Cortical spectral integration and frequency gradient
Fine-grain cortical frequency mapping showed a nonuniform CF gradient in AI. The gradient distribution has two features. First, the mean gradient changes as a function of CF with the steepest gradient <5 kHz. This corresponds to a smaller magnification factor and a relative underrepresentation of those frequencies. A previous study (Merzenich et al. 1975) had also noted a nonuniform frequency representation and interpreted it as a potential overrepresentation of the mid-frequency range. A comparison between the cortical and cochlear frequency gradients (Fig. 8B) suggests that the cortical mid-frequency magnification matches that of the cochlea. As explained in the preceding text, the shallow cortical gradient observed in the high-frequency range is likely influenced by the speaker system. A second aspect of the CF-gradient distribution is that most frequency bands exhibit a wide range of gradients (Fig. 8). For the mid-frequency range, this reflects steeper gradients at the dorsal and ventral poles of isofrequency contours and less ordered CF organization near the borders of the other auditory fields (Fig. 6). However, the AI tonotopic gradient is smooth relative to the other primary field, AAF, that has gross local distortions and even omissions in its CF representation (Imaizumi et al. 2004). The functional interpretation of the observed differences between cortical and estimated cochlear frequency gradients remains difficult. The issue of an under- or overrepresentation of certain frequency regions in AC (Merzenich et al. 1975) depends on a precise estimate of the AC extent, which is hampered by the sulci. Furthermore, the observation of different frequency gradients along the isofrequency domain suggests that a functional interpretation of cortical frequency gradient requires experimental approaches that are more task- and region-specific than provided by a simple RF analysis.
The current findings suggest that the spatial variations in frequency gradient are related to variations in spectral integration across AI. The spectral integration properties of AI neurons are shaped by many factors that affect the excitatory and inhibitory components of cell inputs and the subsequent determination of action potential generation. An unresolved question about the generation of cortical spectral RFs is whether there is a match, or at least proportionality, between a cortical neuron's output bandwidth and the converging thalamic and cortical frequency information. Intracellular studies find that the bandwidths of excitatory and inhibitory inputs are well matched (Tan et al. 2004; Wehr and Zador 2003) and are usually wider than the RF derived from action potentials. The inhibitory contributions are all of cortical origin and predominantly from local interneurons (Prieto et al. 1994). It is reasonable to predict that local excitatory cortical contributions will match the spectral extent of the inhibitory inputs. However, it is not clear how thalamic and cortical spectral bandwidths map onto the cortical output. Cross-correlation studies of spiking activity indicate that the excitatory portions of connected thalamocortical neuron pairs either closely match in CF and bandwidth or only partially overlap, i.e., they can differ in CF (by up to an 1/3 octave) and bandwidth (Miller et al. 2001). The former case suggests cortical inheritance of thalamic spectral tuning, and the latter case supports new RF construction and, thus, a change in spectral integration.
A parsimonious scenario is that the spectral extent of the thalamic and cortical contributions covary: narrow-band neurons link narrow and frequency-matched thalamic and cortical inputs; broad–band neurons join either broadly tuned thalamic and cortical inputs and/or narrowly tuned inputs that are remote in CF. This does not imply, however, that the spectral convergence from thalamic and cortical sources must be identical. It has been proposed that the core-region of cortical RFs may be predominantly supplied by thalamic inputs, whereas flanks may be dominated by cortical input (Metherate et al. 2005). The critical parameters for spectral convergence, in any case, are the CF range and the individual neuronal bandwidth of the thalamic and cortical neurons that contribute to the creation of either narrow- or broad-band cortical modules.
Concluding remarks
We find two interactions between the CF organization and spectral bandwidth modules in cat AI: the spectral integration range and the expression of distinct bandwidth modules are frequency dependent and the tonotopic gradient, at least as a first approximation, covaries with the spectral integration range. These findings allow more refined models of thalamocortical and corticocortical spectral convergence and integration to be developed and tested. This is of specific significance for the question whether AI receives and/or is the source of distinct functional channels and how these might contribute to signal processing in other cortical fields. These findings also suggest modality-specific organizational principles that may help us to delineate general cortical processing features.
|
|
GRANTS |
|---|
|
|
|
ACKNOWLEDGMENTS |
|---|
|
|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: K. Imaizumi, W. M. Keck Center for Integrative Neuroscience, University of California San Francisco, 513 Parnassus Ave., Box 0732, San Francisco, CA 94143-0732 (E-mail: kazuo{at}phy.ucsf.edu)
|
|
REFERENCES |
|---|
|
Bushnell MC, Duncan GH, Hofbauer RK, Ha B, Chen JI, Carrier B. Pain perception: is there a role for primary somatosensory cortex? Proc Natl Acad Sci USA 96: 7705–7709, 1999.
Cheung SW, Bedenbaugh PH, Nagarajan SS, Schreiner CE. Functional organization of squirrel monkey primary auditory cortex: responses to pure tones. J Neurophysiol 85: 1732–1749, 2001.
Eggermont JJ. Between sound and perception: reviewing the search for a neural code. Hear Res 157: 1–42, 2001.[CrossRef][Web of Science][Medline]
Ehret G, Schreiner CE. Frequency resolution and spectral integration (critical band analysis) in single units of the cat primary auditory cortex. J Comp Physiol [A] 181: 635–650, 1997.[CrossRef][Web of Science][Medline]
Greenwood DD. A cochlear frequency-position function for several species—29 years later. J Acoust Soc Am 87: 2592–2605, 1990.[CrossRef][Web of Science][Medline]
He J, Hashikawa T. Connections of the dorsal zone of the cat auditory cortex. J Comp Neurol 400: 334–348, 1998.[CrossRef][Web of Science][Medline]
Huang CL, Winer JA. Auditory thalamocortical projections in the cat: laminar and areal patterns of input. J Comp Neurol 427: 302–331, 2000.[CrossRef][Web of Science][Medline]
Hubel DH, Wiesel TN. Ferrier lecture. Functional architecture of macaque monkey visual cortex. Proc R Soc Lond B Biol Sci 198: 1–59, 1977.[Medline]
Hübener M, Shoham D, Grinvald A, Bonhoeffer T. Spatial relationships among three columnar systems in cat area 17. J Neurosci 17: 9270–9284, 1997.
Imaizumi K, Priebe NJ, Crum PAC, Bedenbaugh PH, Cheung SW, Schreiner CE. Modular functional organization in cat anterior auditory field. J Neurophysiol 92: 444–457, 2004.
Imig TJ, Adrián HO. Binaural columns in the primary field (AI) of cat auditory cortex. Brain Res 138: 241–257, 1977.[CrossRef][Web of Science][Medline]
Imig TJ, Brugge JF. Sources and terminations of callosal axons related to binaural and frequency maps in primary auditory cortex of the cat. J Comp Neurol 182: 637–660, 1978.[CrossRef][Web of Science][Medline]
Inouye T. Visual Disturbances Following Gunshot Wounds of the Cortical Visual Area: Oxford, UK: Oxford Univ. Press, 2000.
Issa NP, Trepel C, Stryker MP. Spatial frequency maps in cat visual cortex. J Neurosci 20: 8504–8514, 2000.
Kelly JB, Sally SL. Organization of auditory cortex in the albino rat: binaural response properties. J Neurophysiol 59: 1756–1769, 1988.
Knight PL. Representation of the cochlear within the anterior auditory field (AAF) of the cat. Brain Res 130: 447–467, 1977.[CrossRef][Web of Science][Medline]
Liberman MC. Auditory-nerve response from cats raised in a low-noise chamber. J Acoust Soc Am 63: 442–455, 1978.[CrossRef][Web of Science][Medline]
Merzenich MM, Knight PL, Roth GL. Representation of cochlea within primary auditory cortex in the cat. J Neurophysiol 38: 231–249, 1975.
Metherate R, Kaur S, Kawai H, Lazar R, Liang K, Rose HJ. Spectral integration in auditory cortex: mechanisms and modulation. Hear Res 206: 146–158, 2005.[CrossRef][Web of Science][Medline]
Middlebrooks JC, Xu L, Furukawa S, Mickey BJ. Location signaling by cortical neurons. In: Integrative Functions in the Mammalian Auditory Pathway, edited by Oertel D, Popper AN, Fay RR. New York: Springer-Verlag, 2002, p. 319–357.
Miller LM, Escabí MA, Read HL, Schreiner CE. Functional convergence of response properties in the auditory thalamocortical system. Neuron 32: 151–160, 2001.[CrossRef][Web of Science][Medline]
Narayan SS, Temchin AN, Recio A, Ruggero MA. Frequency tuning of basilar membrane and auditory nerve fibers in the same cochleae. Science 282: 1882–1884, 1998.
Nelken I. Feature detection by the auditory cortex. In: Integrative Functions in the Mammalian Auditory Pathway, edited by Oertel D, Fay RR, Popper AN. New York: Springer-Verlag, 2002, p. 358–416.
Nienhuys TG, Clark GM. Critical bands following the selective destruction of cochlear inner and outer hair cells. Acta Otolaryngol 88: 350–358, 1979.[Medline]
Noreña AJ, Gourevich B, Aizawa N, Eggermont JJ. Spectrally enhanced acoustic environment disrupts frequency representation in cat auditory cortex. Nat Neurosci 9: 932–939, 2006.[CrossRef][Web of Science][Medline]
Philibert B, Beitel RE, Nagarajan SS, Bonham BH, Schreiner CE, Cheung SW. Functional organization and hemispheric comparison of primary auditory cortex in the common marmoset (Callithrix jacchus). J Comp Neurol 487: 391–406, 2005.[CrossRef][Web of Science][Medline]
Phillips DP, Semple MN, Calford MB, Kitzes LM. Level-dependent representation of stimulus frequency in cat primary auditory cortex. Exp Brain Res 102: 210–226, 1994.[Web of Science][Medline]
Pickles JO. Normal critical bands in the cat. Acta Otolaryngol 80: 245–254, 1975.[Medline]
Prieto JJ, Peterson BA, Winer JA. Morphology and spatial distribution of GABAergic neurons in cat primary auditory cortex (AI). J Comp Neurol 344: 349–382, 1994.[CrossRef][Web of Science][Medline]
Razak KA, Fuzessery ZM. Functional organization of the pallid bat auditory cortex: emphasis on binaural organization. J Neurophysiol 87: 72–86, 2002.
Razak KA, Fuzessery ZM. Neural mechanisms underlying selectivity for the rate and direction of frequency-modulated sweeps in the auditory cortex of the pallid bat. J Neurophysiol 96: 1303–1319, 2006.
Read HL, Winer JA, Schreiner CE. Modular organization of intrinsic connections associated with spectral tuning in cat auditory cortex. Proc Nat Acad Sci USA 98: 8042–8047, 2001.
Reale RA, Imig TJ. Tonotopic organization in auditory cortex of the cat. J Comp Neurol 192: 265–291, 1980.[CrossRef][Web of Science][Medline]
Recanzone GH, Schreiner CE, Sutter ML, Beitel RE, Merzenich MM. Functional organization of spectral receptive fields in the primary auditory cortex of the owl monkey. J Comp Neurol 415: 460–481, 1999.[CrossRef][Web of Science][Medline]
Reser DH, Fishman YI, Arezzo JC, Steinschneider M. Binaural interactions in primary auditory cortex of the awake macaque. Cereb Cortex 10: 574–584, 2000.
Schreiner CE, Cynader MS. Basic functional organization of second auditory cortical field (AII) of the cat. J Neurophysiol 51: 1284–1305, 1984.
Schreiner CE, Mendelson JR. Functional topography of cat primary auditory cortex: Distribution of integrated excitation. J Neurophysiol 64: 1442–1459, 1990.
Schreiner CE, Read HL, Sutter ML. Modular organization of frequency integration in primary auditory cortex. Annu Rev Neurosci 23: 501–529, 2000.[CrossRef][Web of Science][Medline]
Schreiner CE, Sutter ML. Topography of excitatory bandwidth in cat primary auditory cortex: single-neuron versus multiple-neuron recordings. J Neurophysiol 68: 1487–1502, 1992.
Shamma SA, Fleshman JW, Wiser PR, Versnel H. Organization of response areas in ferret primary auditory cortex. J Neurophysiol 69: 367–383, 1993.
Suga N. Auditory neuroethology and speech processing: complex-sound processing by combination-sensitive neurons. In: Auditory Function. Neurobiological Bases of Hearing, edited by Edelman GM, Gall WE, Cowan WM. New York: Wiley, 1988, p. 679–720.
Suga N. Sharpening of frequency tuning by inhibition in the central auditory system: tribute to Yasuji Katsuki. Neurosci Res 21: 287–299, 1995.[CrossRef][Web of Science][Medline]
Sutter ML, Schreiner CE. Physiology and topography of neurons with multipeaked tuning curves in cat primary auditory cortex. J Neurophysiol 65: 1207–1226, 1991.
Sutter ML, Schreiner CE, McLean M, O'Connor KN, Loftus WC. Organization of inhibitory frequency receptive fields in cat primary auditory cortex. J Neurophysiol 82: 2358–2371, 1999.
Swindale NV, Shoham D, Grinvald A, Bonhoeffer T, Hübener M. Visual cortex maps are optimized for uniform coverage. Nat Neurosci 3: 822–826, 2000.[CrossRef][Web of Science][Medline]
Tan AY, Zhang LI, Merzenich MM, Schreiner CE. Tone-evoked excitatory and inhibitory synaptic conductances of primary auditory cortex neurons. J Neurophysiol 92: 630–643, 2004.
Teng C, Imaizumi K, Schreiner CE. Feature map interactions between binaural and spectral processing in primary auditory cortex. Soc Neurosci Abstr 800.1, 2006.
Wallace MN, Rutkowski RG, Palmer AR. Interconnections of auditory areas in the guinea pig neocortex. Exp Brain Res 143: 106–119, 2002.[CrossRef][Web of Science][Medline]
Wehr M, Zador AM. Balanced inhibition underlies tuning and sharpens spike timing in auditory cortex. Nature 426: 442–446, 2003.[CrossRef][Medline]
Yin TCT, May BJ. Acoustic behavior and midbrain function. In: The Inferior Colliculus, edited by Winer JA, Schreiner CE. New York: Springer, 2005, p. 426–458.
Young ED, Davis KA. Circuitry and function of the dorsal cochlear nucleus. In: Integrative Functions in the Mammalian Auditory Pathway, edited by Oertel D, Fay RR, Popper AN. New York: Springer-Verlag, 2002, p. 160–206.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
TOP
ABSTRACT
INTRODUCTION
|, significant difference between the 2 CF ranges.
