Asymmetrical representation of auditory space in human cortex

Transcription

1 available at Research Report Asymmetrical representation of auditory space in human cortex Nelli H. Salminen a,c,, Hannu Tiitinen a,c, Ismo Miettinen a,c, Paavo Alku b, Patrick J.C. May a,c a Department of Biomedical Engineering and Computational Science, Helsinki University of Technology, Finland b Department of Signal Processing and Acoustics, Helsinki University of Technology, Finland c BioMag Laboratory, Hospital District of Helsinki and Uusimaa HUSLAB, Helsinki University Central Hospital, Finland ARTICLE INFO Article history: Accepted 24 September 2009 Available online 30 September 2009 Keywords: Sound source localization Human MEG N1m Stimulus-specific adaptation ABSTRACT Recent single-neuron recordings in monkeys and magnetoencephalography (MEG) data on humans suggest that auditory space is represented in cortex as a population rate code whereby spatial receptive fields are wide and centered at locations to the far left or right of the subject. To explore the details of this code in the human brain, we conducted an MEG study utilizing realistic spatial sound stimuli presented in a stimulus-specific adaptation paradigm. In this paradigm, the spatial selectivity of cortical neurons is measured as the effect the location of a preceding adaptor has on the response to a subsequent probe sound. Two types of stimuli were used: a wideband noise sound and a speech sound. The cortical hemispheres differed in the effects the adaptors had on the response to a probe sound presented in front of the subject. The right-hemispheric responses were attenuated more by an adaptor to the left than by an adaptor to the right of the subject. In contrast, the lefthemispheric responses were similarly affected by adaptors in these two locations. When interpreted in terms of single-neuron spatial receptive fields, these results support a population rate code model where neurons in the right hemisphere are more often tuned to the left than to the right of the perceiver while in the left hemisphere these two neuronal populations are of equal size Elsevier B.V. All rights reserved. 1. Introduction Auditory localization has a unique role in orienting in the environment. While vision is limited to locations in front of the perceiver, audition allows the detection of objects in all directions and even when they are obscured by visual obstacles. Consequently, the auditory system provides crucial spatial information for directing other senses towards interesting objects and events. Recent studies suggest that sound source location in the horizontal plane is represented in the cortex by a population rate code. According to this model, auditory spatial receptive fields are centered either to the left or right of the subject and span the whole hemifield. This neural code has been studied in detail in the monkey with single-neuron recordings (Woods et al., 2006; Werner-Reiss and Groh, 2008). Human psychophysical (Boehnke and Phillips, 1999) and neuroimaging (Salminen et al., in press) results are also consistent with the population rate code but very little is Corresponding author. Department of Biomedical Engineering and Computational Science, Helsinki University of Technology, P. O. Box 2200, FI TKK, Finland. Fax: address: nelli.salminen@tkk.fi (N.H. Salminen) /$ see front matter 2009 Elsevier B.V. All rights reserved. doi: /j.brainres

2 94 BRAIN RESEARCH 1306 (2010) still known about the details of its neuronal implementation. Here, we conducted magnetoencephalography (MEG) measurements to study the differences between the representations of auditory space in the two cortical hemispheres. Single-unit recordings in the monkey show that the implementation of the population rate code is asymmetrical in the two cortical hemispheres (Benson et al., 1981; Ahissar et al., 1992; Woods et al., 2006). Neurons are more often tuned to the hemifield contralateral to the measurement site than to the ipsilateral side. Indications of contralateral preference have also been found in the human cortex. Stronger activity is often found in the hemisphere contralateral to the sound source location (Ungan et al., 2001; Palomäki et al., 2005; Krumbholz et al., 2005, 2007) and this could potentially reflect a larger number of neurons being tuned to the contralateral than to the ipsilateral hemifield. The right and the left hemisphere seem to differ in the magnitude of the contralateral preference although this issue is not entirely clear: some studies show a stronger contralateral preference in the right than in the left hemisphere (Palomäki et al., 2002, 2005; Tiitinen et al., 2006) while others suggest that the asymmetry is larger in the left hemisphere (Ungan et al., 2001; Krumbholz et al., 2005, 2007). Thus, it remains unresolved what the relative sizes of the contralaterally and ipsilaterally tuned populations are in the human cortex. Psychophysical studies have revealed several sources of location information embedded in sound (Middlebrooks and Green, 1991). At low sound frequencies, the most prominent cue is the interaural time difference (ITD) that results from the sound reaching one ear before the other and occurs both at the onset and along the whole duration of the sound. At higher frequencies, the head of the listener casts a shadow on the sound and thus creates an interaural level difference (ILD). Also, at higher frequencies, the pinnae, the head, and the body of the listener alter the frequency spectrum of the sound in a direction-specific manner. These cues do not occur as independent entities in the sound waveform. Instead, they are imposed on sound signals that already have complex spectrotemporal structure and, consequently, the availability and usefulness of different types of localization cues diverges from one sound source to another. For instance, while a wide-band noise sound includes all of the localization cues, a speech sound has less energy in the high frequencies and, thus, carries less of the spectral cues and has a smaller ILD. How this is reflected in the cortical representations of sound source location is not known. The N1m response, occurring in the event-related field at around 100 ms after sound onset, and its electrical counterpart N1 have proven to be useful measures of spatial sensitivity in the human auditory cortex. The N1m is generated in multiple secondary (belt and parabelt) areas of auditory cortex (for a review, see May and Tiitinen, in press) and it varies in amplitude according to sound source location. It is maximal for sources contralateral to the hemisphere from which it is measured and minimal for ipsilateral sources (Palomäki et al., 2005). The N1 and N1m responses, when obtained in a stimulus-specific adaptation paradigm, can provide information on the spatial selectivity of auditory cortical neurons (Butler, 1972; Salminen et al., in press). In this paradigm, sounds are presented in adaptor-probe pairs and the effect of the preceding adaptor on the response to the following probe is measured (Fig. 1). When the adaptor is presented at the same location as the probe, the two sounds activate the same population of spatially selective neurons and, consequently, the attenuation of the amplitude of the N1m response is maximal. However, when the adaptor is presented from a different location than the probe, the neurons activated by the probe but not by the adaptor are presumably not influenced by the preceding adaptor presentation and, thus, respond to the probe strongly. This selectivity, then, leads to an increase in the amplitude of the N1m response. The aim of the present study was to compare the spatial selectivity between the cortical hemispheres and to estimate the relative sizes of the contralaterally and ipsilaterally tuned neuronal populations in each hemisphere. To this end, we presented spatial sound stimuli, individually prepared for each subject, in a stimulus-specific adaptation paradigm and measured the attenuation of the N1m response. To facilitate comparisons between the hemispheres, the sound sources were situated symmetrically with respect to the midline. The probe sound was always directly in front of the subject and adaptors were presented in the left and right hemifields. The Fig. 1 Illustration of the stimulus-specific adaptation paradigm. (A) Two sounds, an adaptor and a probe are sequentially presented and the brain responses to the probe are measured. When the adaptor and the probe are presented from the same location directly in front of the subject, adaptation is maximal and responses are small. (B) The adaptor is then presented from a location 45 to the right from the probe. Assuming that the neuronal population giving rise to the response is selective to sound source location, this leads to a less attenuated response to the probe. (C) The largest responses occur when no adaptor sound is presented. (D) The difference between the response amplitudes measured in these conditions can be used as a measure of neuronal selectivity to sound source location.

3 95 experiment was designed to measure the relative numbers of contralaterally and ipsilaterally tuned neurons. Assuming that the two populations are of equal size, the adaptors in the left and right should be equally effective in attenuating the response to the probe. However, if one of the populations is larger, asymmetrical attenuation of the N1m should occur. We also aimed to study how the neuronal representation of sound source location is formed for different types of sounds. Therefore, spatial selectivity was tested with both a wideband noise sound and a vowel sound. 2. Results The N1m responses were measured from the left and right hemisphere to a probe sound either a noise or a vowel stimulus presented directly in front of the subject in the context of adaptors at three different locations or without an intervening adaptor (Fig. 2). In both hemispheres, the vowel sound led to responses with approximately twice the amplitude of those elicited by the noise sound (main effect of stimulus type: F[1,10]=53.3, p<0.001). In the right hemisphere, the spatial sound stimuli elicited N1m responses with average amplitudes of 30.1 ft/cm and 57.1 ft/cm for the noise and the vowel sound, respectively. For the left-hemispheric N1m, these amplitudes were 26.3 ft/cm and 55.5 ft/cm. Additionally, the N1m peak amplitudes were slightly larger in the right than in the left hemisphere, although this difference did not reach statistical significance (main effect of hemisphere: F[1,10] = 0.15, p=n.s.). The amplitude of the N1m depended on the adaptor condition (main effect of adaptor: F[3,30]=51.3, p<0.001). In both hemispheres and for both stimulus types, the smallest responses to the probe were measured when the adaptor was at the same location as the probe, and the largest responses occurred when no adaptors were presented (Fig. 2). The amplitude of the N1m response measured in the other conditions fell between these two extremes. The effect of the adaptor on the N1m response amplitude depended on the hemisphere and the stimulus type (threeway interaction: F[3,30] =2.9, p=0.05). The response amplitudes in the no-adaptor condition were larger for the speech stimulus in the left than in the right hemisphere (83.0 and 76.1 ft/cm, for the left and right hemisphere, respectively, p<0.01). For the noise sound, no such asymmetry was found (41.0 and 43.5 ft/cm, p=n.s.). Thus, the overall dynamic range of the N1m response amplitude varied across hemispheres and stimulus types. To analyze the location-specific adaptation of the N1m response independent of variations in the overall level of activity, the reduction in attenuation caused by the adaptors being to the left and to the right instead of being directly in front was expressed relative to the overall dynamic range of the N1m responses to the same stimulus type in the same hemisphere (see Fig. 1). Further analyses were all performed on these relative measures of location-specificity of the response. When the N1m amplitude was analyzed as a measure relative to the dynamic range of the response, a significant interaction between the hemisphere and the adaptor location was found (Fig. 3; F[1,10] =11.2, p<0.01). In the right hemisphere, Fig. 2 Sound source locations and averaged event-related fields from a representative measurement channel. (A) Stimulus-specific adaptation of the N1m response was measured in four conditions. The probe sound presented always directly in front (0 ) was coupled with adaptors at the same location (0 ), to the left ( 45 ), or to the right (+45 ) of the subject, or no adaptor was presented. (B) Event-related fields to the probe presented in four adaptor conditions and for two stimulus types (noise and vowel) were measured and averaged over eleven subjects. In both hemispheres and for both stimulus types, the smallest response amplitudes were measured when the adaptor was at the same location as the probe and the largest when no adaptor was presented. With the adaptor to the left or right of the subject, the response amplitudes fell between these two extremes. Thus, location-specific adaptation was found in both hemispheres and for both noise and vowel stimuli. Fig. 3 The N1m amplitude expressed as relative to the overall dynamic range. The error bars depict the standard error of the mean. In the left hemisphere, the adaptors to the left and to the right of the subject led to similar increases in the N1m amplitude. However, in the right hemisphere, the increase was much larger when the adaptor was to the right than when it was to the left.

4 96 BRAIN RESEARCH 1306 (2010) the adaptor in the left hemifield caused a larger attenuation of the N1m response than did the adaptor in the right hemifield (p<0.01). This occurred for both noise and vowel stimuli but the effect was larger for the noise stimulus. Thus, in the right hemisphere, the adaptor in the contralateral hemifield caused stronger attenuation of the N1m response than did the adaptor in the ipsilateral hemifield. For the left-hemispheric N1m responses, the effect the adaptor had on the response to the probe did not depend on whether the adaptor was to the left or to the right (p=n.s). The differences found between the left and right adaptor conditions were small and inconsistent across the stimulus types. The N1m response peaked, on the average, at the latency of 104 ms from stimulus onset. No effect of hemisphere (F[1,10] = 0.10, p=n.s.) or stimulus type (F[1,10]=1.62, p=n.s.) was found. The peak latency, however, depended on the presence of the adaptor stimulus (F[1,10] =6.18, p<0.01). In the no-adaptor condition the N1m peaked at 100 ms while in the other conditions the peak latency was 105 ms (p<0.05). 3. Discussion We conducted an MEG experiment utilizing realistic, individually prepared spatial sound stimuli in a stimulus-specific adaptation paradigm to study the cortical encoding of auditory space. Our aim was, first, to compare the spatial selectivity of the two cortical hemispheres and, second, to elucidate whether the spatial selectivity to speech sounds differs from that of the noise sounds. We measured N1m responses elicited by a probe sound presented directly in front of the subject in the context of adaptors either to the left or to the right. A notable difference between the hemispheres was found in the effect the adaptor location had on the response to the probe. In the right hemisphere, the attenuation of the N1m response was much stronger when the adaptor was to the left than when it was to the right. In the left hemisphere, no such asymmetry was found. In general, the response amplitudes were larger for the vowels than for the noise sounds, with the left hemisphere generating especially prominent responses to the vowels. However, the right-hemispheric measures of spatial selectivity obtained with noise stimulation were clearly more prominent than those gained with the vowel stimuli. Thus, the right hemisphere appears to be especially selective to spatial information, and this selectivity is greater for noise sounds than for speech sounds. The N1m response elicited by the probe in front probably reflects the combined activity of left- and right-tuned populations similar to those found in monkey auditory cortex (Benson et al., 1981; Ahissar et al., 1992; Woods et al., 2006; Werner-Reiss and Groh, 2008). When the adaptor is presented in the left or in the right hemifield, it attenuates mainly the activity of the population tuned to that hemifield while the population tuned to the opposite hemifield remains largely unaffected (Fig. 4). If the two populations are of equal size, the N1m amplitude is affected similarly by the adaptors in the left and right hemifields (Fig. 4B, top). However, when more neurons are tuned to locations in the left than to those in the right hemifield, the adaptor to the left attenuates a larger number of neurons than the adaptor to the right (Fig. 4B, bottom). This would be reflected as small responses when the adaptor is in the left and large ones when the adaptor is in the right hemifield. This pattern was, in fact, realized by the righthemispheric N1m responses measured here. Therefore, the present results suggest that, in the right hemisphere, more neurons are tuned to the left than to the right hemifield. However, in the left hemisphere, there seems to be little difference between the sizes of these two populations. Fig. 4 Interpretation of the results based on single-neuron spatial receptive fields. (A) The majority of auditory spatial receptive fields found in the cortex are wide and centered at locations either to the left or to the right of the subject. The top figure represents a case where the same number of neurons is tuned to each direction. Often, however, the majority of the neurons are tuned to contralateral locations. The bottom figure represents a hypothetical right hemisphere where neurons tuned to the left outnumber those tuned to the right. (B) The N1m response reflects the compound activity of the left- and right-tuned populations. When the two populations are of equal size they contribute similarly to the N1m response elicited by a probe sound in front (0 ). Thus, the effects of the adaptors to the left ( 45 ) and right (+45 ) of the subject are similar (top). This corresponds to the left-hemispheric results obtained in this study. However, when more neurons are tuned to the left than to the right of the subject, the left-tuned neurons contribute more to the N1m response than the right-tuned neurons. Consequently, the attenuation caused by the adaptor to the left is stronger than that caused by the adaptor to the right (bottom). This is consistent with our right-hemispheric data.

5 97 The above interpretation in terms of single-neuron spatial receptive fields is consistent with earlier MEG observations (Palomäki et al., 2002, 2005; Tiitinen et al., 2006). Previous studies have shown that the auditory cortices respond more strongly to sounds originating from contralateral directions than to those located ipsilaterally. For instance, when sounds are presented from various horizontal locations, the N1m response of each hemisphere has the largest amplitude for contralateral and the smallest amplitude for ipsilateral sound sources. By considering that the N1m represents the compound activity of the neurons of auditory cortex (May and Tiitinen, in press), its amplitude variation would be consistent with the activity of contralaterally tuned neurons, which therefore probably outnumber ipsilaterally tuned neurons. Further, the variation of the N1m response found in previous studies is much weaker in the left than in the right hemisphere (Palomäki et al., 2005). This could be due to the contralaterally and ipsilaterally tuned populations being of nearly equal size in the left hemisphere. In this case, the location-dependent variations of the two populations cancel each other out leading to less variation in the amplitude of the N1m response. Alternatively, the weaker variation of the N1m response depending on sound source location could be explained by a smaller number of spatially selective neurons in the left than in the right hemisphere. The current results, however, clearly demonstrate that the left-hemispheric responses are selective to sound source location. This shows that even though the right hemisphere is specialized in spatial processing (Griffiths et al., 1998; Baumgart et al., 1999; Zatorre and Penhune, 2001; Zatorre et al., 2002; Warren and Griffiths, 2003; Deouell et al., 2007) both cortical hemispheres represent auditory space with location-selective neurons. While the current results suggest a stronger contralateral preference in the right hemisphere, a number of previous studies have demonstrated similar preference in the two hemispheres or even a larger asymmetry in the left hemisphere (Woldorff et al., 1999; Ungan et al., 2001; Jäncke et al., 2002; Krumbholz et al., 2005, 2007; Schönwiesner et al., 2007; Getzmann, 2009). This discrepancy may arise from differences in the sound stimulation. First, larger contralateral preference in the left than in the right hemisphere has been found in studies utilizing the ILD or ITD cue alone while, in the present study, realistic spatial sounds were used. This, however, is unlikely to account fully for the diverging results as the stimuli in the present study contained prominent ITD and ILD cues also. Second, in many of the studies showing strong lefthemispheric contralateral preference (Ungan et al., 2001; Krumbholz et al., 2005, 2007; Getzmann, 2009), the sound stimuli contained changes in ITD or ILD in an ongoing sound. In contrast, the spatial cues of the current study were static. Thus, the discrepancy between the past and present results could reflect differences between sensitivity to stationary and to moving sound sources. In the mammalian subcortical structures, the spatial receptive fields are strongly influenced by sound source movement (Spitzer and Semple, 1991; McAlpine et al., 2000) and this could be the case for the human auditory cortex also. Finally, in studies where a lefthemispheric contralateral preference has been found with stationary stimuli, the sounds were presented monaurally (Pantev et al., 1986; Woldorff et al., 1999; Jäncke et al., 2002; Schönwiesner et al., 2007). These findings are probably related to the distribution of monaural neurons activated exclusively by left- or right-ear stimulation showing that those preferring the contralateral ear are in the majority in each cortical hemisphere. Our findings, in contrast, were obtained with binaural stimulation and therefore, are likely to reflect the activity of binaural neurons sensitive to the interaural difference cues. The distribution of these neurons seems to be more balanced in the left than in the right hemisphere. There were several differences between the N1m responses to the vowel and the noise sounds. Most conspicuously, the responses measured for the vowel sounds were of larger amplitude, and this difference was especially clear in the left hemisphere. This may reflect a left-hemispheric specialization of speech processing which is apparent already at the level of isolated vowel sounds. Then again, the locationdependent effects (observed in the right hemisphere) were more prominent for the noise sounds than for the vowel sounds. This difference could be explained by the spatial cues that the sounds carry. The wide-band noise stimulus contains all the localization cues while in the case of the vowel sound the ILD and spectral cues are weaker and less useful (Middlebrooks and Green, 1991). The location-specific adaptation measured with speech and noise stimuli followed, nonetheless, the same pattern of variation. This suggests that a similar neuronal representation of sound location was arrived at independently of the bandwidth of the sound stimulus. However, it is also possible that top-down effects contributed to the stronger lateralization of location processing for the noise sounds than for the speech sounds. Such top-down effects in the processing of location information is indicated by the results of Schönwiesner et al. (2007) who found that hemispheric lateralization of responses to monaural sounds is not stimulus-specific but, rather, can be modulated in a topdown manner by the context that the stimuli are presented in. Systematic studies on the contributions of different localization cues in various contexts on the cortical representation of auditory space will be needed to understand how the neural code builds on these diverse sources of acoustical information. 4. Experimental procedures 4.1. Subjects Fourteen right-handed subjects (mean age 27, s.t.d. 6 years, 7 females) took part in the experiment. The measurements were performed with the written informed consent of the subject and under the approval of the Ethical Committee of Helsinki University Central Hospital. During the experiment, the subjects were under instruction to focus on a self-selected silent film and to ignore the auditory stimulation. The data of three subjects were discarded due to a poor signal-to-noise ratio Spatial sound stimuli and presentation Binaural recordings were performed in a highly controlled manner to create individual spatial sound stimuli for each subject. The subject was seated in the center of a standardized

6 98 BRAIN RESEARCH 1306 (2010) listening room and was surrounded by a constellation of loudspeakers. Each loudspeaker was at a distance of 1.3 m from the subject's head and at a height of 1.2 m (Fig. 2). One loudspeaker was directly in front of the subject (0 ) and the second and the third speakers were 45 to the left and the right, respectively. The vertical position of the subject was adjusted so that his/her ear canals were always at the same distance of 1.2 m from the floor. White noise and vowel signals of a 200-ms duration were presented sequentially from each loudspeaker and recorded with miniature microphones placed at the entrances of the ear canals of the subject. For the vowel sound, a Finnish /a/ spoken by a male speaker was used. During the MEG measurements, the stimuli were played back to the subject through a custom-made tube-phone system with a flat amplitude response up to 10 khz. The sound level was adjusted to 75 db SPL (A) for the noise and vowel stimuli at 0. The level of the stimuli corresponding to sound locations at 45 and +45 was scaled accordingly. In the stimulus-specific adaptation paradigm sounds were presented in alternating pairs of an adaptor and a probe stimulus with an interstimulus interval of 1 s. Thus, in all stimulus blocks a probe sound originating from directly in front of the subject (0 ) occurred at 2-s intervals. In each stimulus block, the adaptor sound was presented either from the same location as the probe (0 ), or from the left ( 45 ) or right (+45 ) direction. Additionally, a control condition with no intervening adaptors was included. Thus, there was a total of four adaptor conditions resulting in altogether eight stimulus blocks whose presentation order was counterbalanced across subjects MEG data acquisition Data was recorded with a 306-channel whole-head MEG device (Vectorview 4-D, Neuromag Oy, Finland) with a passband of Hz. The event-related responses were averaged online over a time period from 100 ms before until 400 ms after stimulus onset. Eye-movements were monitored with electrodes, and epochs with absolute deviations larger than 150 μv were automatically discarded. A minimum of 150 repetitions were collected for each stimulus condition. The averaged responses were bandpass filtered at 1 30 Hz and baseline-corrected with respect to a 100-ms prestimulus interval. The N1m response amplitude was quantified from a set of planar gradiometer channels above the left and right temporal lobes. For each subject and hemisphere separately, the three channel pairs showing the largest response amplitudes were chosen for further analyses. The N1m response was identified as the amplitude peak in an 80- to 150-ms poststimulus window in the vector sum averaged over these three channel pairs Statistical analyses The amplitudes of the N1m responses were expressed in two ways. First, the absolute amplitude was obtained for each adaptor condition. Second, to provide a measure of locationspecific adaptation relative to the entire dynamic range of the response, the N1m amplitudes measured with the adaptor at ±45 were expressed as the amplitude increase from the adaptor being at 0 and this was divided by the dynamic range of the variation (see Fig. 1D). The dynamic range was evaluated separately for each subject, stimulus type, and hemisphere. This relative measure of location-specific adaptation was obtained to facilitate comparisons of spatial selectivity across the hemispheres and stimulus types independently of the differences in the absolute levels of activity. The absolute and relative N1m amplitudes and the N1m latencies were submitted to repeated-measures ANOVAs. In all analyses, the repeating factors were hemisphere, stimulus type, and adaptor condition. Newman Keuls post-hoc tests were performed when appropriate. Acknowledgments This study was supported by the Academy of Finland (Project Nos , , and ). REFERENCES Ahissar, M.V., Ahissar, E., Bergman, H., Vaadia, E., Encoding of sound-source location and movement: activity of single neurons and interactions between adjacent neurons in the monkey auditory cortex. J. Neurophysiol. 67, Baumgart, F., Gaschler-Markefski, B., Woldorff, M.G., Heinze, H.J., Scheich, H., A movement-sensitive area in the auditory cortex. Nature 400, Benson, D.A., Hienz, R.D., Goldstein, M.H., Single-unit activity in the auditory cortex of monkeys actively localizing sound sources: spatial tuning and behavioral dependency. Brain Res. 219, Boehnke, S.E., Phillips, D.P., Azimuthal tuning of human perceptual channels for sound location. J. Acoust. Soc. Am. 106, Butler, R.A., The influence of spatial separation of sound sources on the auditory evoked response. Neuropsychologia 10, Deouell, L.Y., Heller, A.S., Malach, R., D'Esposito, M., Knight, R.T., Cerebral responses to change in spatial location of unattended sounds. Neuron 55, Getzmann, S., Effect of auditory motion velocity on reaction time and cortical processes. Neuropsychologia 47, Griffiths, T.D., Rees, G., Rees, A., Green, G.G.R., Witton, C., Rowe, D., Büchel, C., Turner, R., Frackowiak, R.S.J., Right parietal cortex is involved in the perception of sound movement in humans. Nat. Neurosci. 1, Jäncke, L., Wüstenberg, T., Schulze, K., Heinze, H.J., Asymmetric hemidynamic responses of the human auditory cortex to monaural and binaural stimulation. Hear. Res. 170, Krumbholz, K., Schönwiesner, M., von Cramon, D.Y., Rübsamen, R., Shah, N.J., Zilles, K., Fink, G.R., Representation of interaural temporal information from left and right auditory space in the human planum temporale and inferior parietal lobule. Cereb. Cortex 15, Krumbholz, K., Hewson-Stoate, N., Schönwiesner, M., Cortical responses to auditory motion suggests an asymmetry in the reliance on inter-hemispheric connections between the left and right auditory cortices. J. Neurophysiol. 97, May, P.J.C., Tiitinen, H., in press. Mismatch negativity (MMN), the deviance-elicited auditory deflection, explained. Psychophysiology.

7 99 McAlpine, D., Jiang, D., Shackleton, T.M., Palmer, A.R., Responses of neurons in the inferior colliculus to dynamic interaural phase cues: evidence for a mechanism of binaural adaptation. J. Neurophysiol. 83, Middlebrooks, J.C., Green, D.M., Sound localization by human listeners. Annu. Rev. Psychol. 42, Palomäki, K.J., Tiitinen, H., Mäkinen, V., May, P.J.C., Alku, P., Cortical processing of speech sounds and their analogues in a spatial auditory environment. Cogn. Brain Res. 14, Palomäki, K.J., Tiitinen, H., Mäkinen, V., May, P.J.C., Alku, P., Spatial processing in human auditory cortex: the effects of 3D, ITD, and ILD stimulation techniques. Cogn. Brain Res. 24, Pantev, C., Lütkenhöner, B., Hoke, M., Lehnertz, K., Comparison between simultaneously recorded auditory-evoked magnetic fields and potentials elicited by ipsilateral, contralateral and binaural tone burst stimulation. Audiology 25, Salminen, N.H., May, P.J.C, Alku, P., Tiitinen, H., in press. A population rate code of auditory space in the human cortex. PLoS ONE. Schönwiesner, M., Krumbholz, K., Rübsamen, R., Fink, G.R., von Cramon, D.Y., Hemispheric asymmetry for auditory processing in the human auditory brain stem, thalamus, and cortex. Cereb. Cortex 17, Spitzer, M.W., Semple, M.N., Interaural phase coding in auditory midbrain: Influence of dynamic stimulus features. Science 254, Tiitinen, H., Salminen, N.H., Palomäki, K.J., Mäkinen, V.T., Alku, P., May, P.J.C., Neuromagnetic recordings reveal the temporal dynamics of auditory spatial processing in the human cortex. Neurosci. Lett. 396, Ungan, P., Yagcioglu, S., Goksoy, C., Differences between the N1 waves of the responses to interaural time and intensity disparities: scalp topography and dipole sources. Clin. Neurophys. 112, Warren, J.D., Griffiths, T.D., Distinct mechanisms for processing spatial sequences and pitch sequence in the human auditory brain. J. Neurosci. 23, Werner-Reiss, U., Groh, J.M., A rate code for sound azimuth in monkey auditory cortex: Implications for human neuroimaging studies. J. Neurosci. 28, Woldorff, M.G., Tempelmann, C., Fell, J., Tegeler, C., Gaschler-Markefski, B., Hinrichs, H., Heinze, H.-J., Scheich, H., Lateralized auditory spatial perception and the contralaterality of cortical processing as studied with functional magnetic resonance imaging and magnetoencephalography. Hum. Brain Mapp. 7, Woods, T.M., Lopez, S.E., Long, J.H., Rahman, J.E., Recanzone, G.H., Effects of stimulus azimuth and intensity on the single-neuron activity in the auditory cortex of the alert macaque monkey. J. Neurophysiol. 96, Zatorre, R.J., Penhune, V.B., Spatial localization after excision of human auditory cortex. J. Neurosci. 21, Zatorre, R.J., Bouffard, M., Ahad, P., Belin, P., Where is where in the human auditory cortex? Nat. Neurosci. 5,