Abstract

Different types of generation mechanisms of 40-Hz auditory steady-state response (ASSR) were investigated using diotic and dichotic stimulation with 500- and 540-Hz pure tones of 1.0-s duration and 2.0-s stimulus onset asynchrony. When the sum of both tones was presented to both ears simultaneously, they interacted at cochlear level and resulted in perception of a 40-Hz beat termed “peripheral beat.” Dichotic presentation of the 500-Hz tone to one ear and the 540-Hz tone to the other one resulted in beat perception as the effect of central interaction, most likely in the superior olivary nuclei and was termed “central beat.” ASSR and transient N1m responses were found in the averaged 151-channel whole-head magnetoencephalographic recordings under both stimulus conditions and were modeled with single spatiotemporal equivalent current dipoles in both hemispheres. The ASSR sources in both conditions were more anterior, more inferior, and more medial compared with N1m sources. Right hemispheric lateralization of the magnetic field strength was found for the ASSR in both stimulus conditions. Although the central and peripheral beat interacted at different levels of the auditory system, the initial responses were projected along the afferent auditory pathway and activated common cortical sources.

Introduction

The 40-Hz auditory steady-state response (ASSR) has been reported in a number of studies since first being demonstrated in electroencephalographic (EEG) recordings from the human brain by Galambos et al. (1981). He demonstrated the sinusoidal waveform of the ASSR with a periodicity corresponding to the stimulation rate and maximum amplitude at 40 Hz. Commonly used auditory stimuli for eliciting ASSR are amplitude-modulated (AM) tones, which show periodical amplitude fluctuations at the modulation frequency. The corresponding ASSR follows the AM envelope of the stimulus (Lins and Picton 1995; John et al. 1998; Engelien et al. 2000). The ability of the central auditory system to respond with steady oscillation at the modulation frequency of the AM sound was explained by Regan D and Regan MP (1988) as demodulation at peripheral level. By means of monaurally presented tones modulated simultaneously with 2 different frequencies of 38 and 40 Hz, we have shown that even a sound with such a complex stimulus envelope was demodulated in the cochlea and that this resulted in 2 distinct ASSRs at cortical level (Draganova et al. 2002). In a follow-up study using 2 AM sounds of different carrier frequencies, we demonstrated strong interference between both sounds and tuning curves of ASSR around 40 Hz, which were clearly distinct from tuning curves measured with evoked responses from various levels of the auditory pathway. These results suggest an ASSR generation in the central auditory pathway (Ross et al. 2003), whereas the envelope extraction through demodulation takes place at cochlear level.

Psychophysical investigations have shown that the physical interaction of 2 pure tones (<1500 Hz) with slightly different frequencies produce sensation of a beat at the difference frequency, which has been termed “monaural beating” (Oster 1973). The dichotic presentation of 2 pure tones with different frequencies causes a beat perception, which results from interaction of bilateral inputs at higher levels of the ascending auditory pathway (Tobias 1963; Rutschmann and Rubinstein 1965), and has been termed “binaural beat.” As the electrophysiological correlate of binaural beat perception, Schwarz and Taylor (2005) recorded 40 Hz ASSR with EEG for the low carrier frequencies of 400 Hz but not for higher frequencies around 3200 Hz.

The present study investigated 2 different mechanisms of the 40-Hz ASSR generation using 2 stimulus conditions: 1) 2 dichotically presented pure tones of 500 and 540 Hz (central beat and its ASSR correlate) and 2) the sum of both tones presented diotically (peripheral beat and its ASSR correlate). The corresponding cortical sources of the ASSR under these 2 conditions were estimated by means of magnetoencephalography (MEG). The leading hypothesis was that even though diotic and dichotic stimuli interact at different places along the auditory pathway, the ASSR may result from a commonly activated cortical network. The study used comparison of waveforms, amplitude spectra, and source locations for both responses in order to demonstrate differences and commonalities of ASSR related to beat perception based on different physiological mechanisms. The developed protocol could be used for further investigation of the central beat mechanism and its projection as ASSR in the auditory cortex, allowing us to assess interaction between crossing auditory pathways.

Methods

Test Subjects

Eleven right-handed subjects (2 females) between 28 and 45 years of age participated in this study. None of them had a history of otological or neurological disorders. Normal audiological status, defined as air conduction threshold of no more than 10-dB hearing level between 250 and 4000 Hz, was verified by pure tone audiometry. Informed consent was obtained from all subjects as approved by the Research Ethics Board at Baycrest Centre, University of Toronto.

Stimulation

The auditory stimuli were 500- and 540-Hz tones, which were presented separately to the left and right ears in the first condition and as the sum of both tones to both ears in the second condition. Figure 1A depicts how the 40-Hz difference between the dichotically presented tones results in periodically increasing and decreasing interaural phase difference. In the sum of both tones, the 40-Hz difference resulted in periodically fluctuating amplitude (Fig. 1B). Initially, and every 25 ms, the tones were in phase and the sum reached maximum amplitude. After 12.5 ms, the tones were of opposite polarity and canceled each other out. The resulting envelope is of the form A = 2·abs(cos(πΔft)) and is shown in the Figure 1B, compared with sinusoidal amplitude modulation with the envelope A = 1+cos(2πfmt), in which the modulation frequency fm equals the difference frequency Δf. In the dichotic condition, the effective stimulus is an interaural phase difference, which oscillates at 40 Hz and evokes the percept of a binaural beat. By means of careful stimulation design, we rejected the possibility of contribution from acoustical crossover from one ear to the other. The interaural attenuation of insert earphones used in our study is typically greater than 75 dB at 500 Hz (Killion 1985). Given a smaller interaural attenuation of only 45 dB, a tone presented to one ear at an intensity of 60 dB would have caused a contralateral intensity of 15 dB. The summation of the cross talk to the ipsilateral sound could have resulted in 40-Hz fluctuations in the envelope with a modulation depth on the order of the interaural attenuation. In a previous study, we found that the ASSR to 500-Hz AM tones at 60 dBSL (sensation level) was below the detection threshold when the modulation depth was below 30 dB. Thus, even an interaural attenuation of 45 dB was sufficient to exclude the possibility of ASSR generation by interaural cross talk. Therefore, we can conclude that the recorded 40-Hz ASSR during dichotic stimulation was generated after central processing of the binaural information.

Figure 1.

Waveforms of the auditory stimuli. A: The 500- and 540-Hz tones presented to the left and right ears in the dichotic stimulation are in phase at t = 0, of opposite phase at t = 12.5 ms (arrows), and again in phase after 25 ms, the period of 40-Hz oscillation. B: The sum of both tones presented to both ears in the diotic stimulation shows envelope fluctuations with 40-Hz periodicity.

Figure 1.

Waveforms of the auditory stimuli. A: The 500- and 540-Hz tones presented to the left and right ears in the dichotic stimulation are in phase at t = 0, of opposite phase at t = 12.5 ms (arrows), and again in phase after 25 ms, the period of 40-Hz oscillation. B: The sum of both tones presented to both ears in the diotic stimulation shows envelope fluctuations with 40-Hz periodicity.

The amplitude fluctuation in the sum of both tones used for diotic stimulation can be perceived even when monaurally presented because the beats exist in the stimulus and can be registered at cochlear level.

All stimuli, each of 1-s duration including 10-ms rise and decay times, were presented repeatedly with 2-s stimulus onset asynchrony at 60 dBSL. The dichotic and diotic stimuli were presented in 2 successive sessions (1 week apart) with session order randomized between subjects. The MEG was recorded for each stimulus condition in six 512-s blocks, each containing 256 stimulus repetitions. The acoustic stimuli were delivered through a nonmagnetic and echo-free acoustic transmission system to silicon earpieces placed into the ear canals.

Data Acquisition

MEG recordings were performed using a 151-channel whole-head neuromagnetometer system (VSM Medtech Inc, Coquitlam, British Columbia, Canada) in a magnetically shielded room. The centers of the detection coils of this system are spaced 31 mm apart in a helmet-shaped array. The sensors were configured as first-order gradiometers with a baseline of 50 mm (Vrba and Robinson 2001). The spectral density of the intrinsic noise of each magnetic channel was below forumla for frequencies above 1 Hz.

The subjects were seated in upright position as comfortably as possible while ensuring that they did not move during the measurement. The subject's head position was determined at the beginning and end of each recording block by means of 3 localization coils fixed to the nasion and the entrances of both ear canals. Subjects were instructed not to move and to stay in a relaxed waking state during the measurement. Alertness and compliance were verified by video monitoring. To control for confounding changes in attention and vigilance, subjects watched a soundless movie of their choice.

Data Analysis

The magnetic field signals were low-pass filtered at 100 Hz, sampled at a rate of 312.5 Hz, continuously recorded in 512-s blocks, and stored for further analysis. For each experimental block, stimulus-related epochs of magnetic field data of 1.4-s duration, including a 200-ms pre- and 1200-ms poststimulus interval, were averaged. Epochs contaminated by eye blink artifacts containing field amplitudes of more than 3 pT in any channel were automatically rejected from the averaging procedure. Although the analysis of the data was focused on the ASSRs, the transient N1m component of the auditory evoked field was also analyzed. Its source location was used as an individual internal cortical marker for comparison with the ASSR source localizations. Complex demodulation at 40 Hz was applied to the averaged magnetic field data in order to extract the 40-Hz ASSR and to reduce the noise prior to the dipole-fit procedure. The procedure calculated the discrete Fourier coefficients (real and imaginary part) at the frequency of interest. The coefficients were rotated onto a common phase for each hemisphere separately with the constraint of best explaining the field variance. Single equivalent current dipoles (ECDs) in both hemispheres were fitted to the resulting 40-Hz amplitudes. Additionally, the averaged magnetic field waveforms were 30-Hz low-pass filtered for analysis of the transient N1m response. Same source analysis based on the model of a single spatiotemporal ECD in a spherical volume conductor for both hemispheres was applied to ASSR and N1m. An ECD for one hemisphere was fitted simultaneously with an ECD for the other hemisphere defined by their dipole moments, orientation, and spatial coordinates. The dipolar source for N1m response was fitted to the peak around the local maximum (approximately 30 ms) of the global field power of the filtered magnetic field data.

The dipole location was determined in a head-based Cartesian coordinate system with the origin at the midpoint of the mediolateral axis (y axis), which joined the center points of the entrances to the ear canals (positive toward the left ear). The posterior–anterior axis (x axis) was oriented from the origin to the nasion (positive toward the nasion), and the inferior–superior axis (z axis) was perpendicular to the xy plane (positive toward the vertex). Source locations fulfilling the following anatomical considerations characterizing the human auditory cortex area were included for further analysis: anterior–posterior value (x) within ±3 cm, medial–lateral value (y—distance from the midsagittal plane) greater than 2.5 cm. Additionally, the statistical consideration of goodness of fit of the dipolar source greater than 85% was imposed. Median values of x, y, and z coordinates of the ECDs and of the angles of the dipole orientation were calculated across all blocks for the same stimulus condition for each subject. The median values of the source coordinates and orientations were averaged across all subjects in each condition. They were used as a reference for estimation of the source wave (the time course of the dipole moment) by the source-space projection method (Tesche et al. 1995). The source-space projection estimates the activity in a certain brain area by a linear combination of the measured field at the 151 sensor positions outside the head. The result is 2 single time series of magnetic dipole moments for the left and right hemispheres. These time series reach a maximum only for a typical dipolar magnetic field pattern of a single current source in an a priori specified brain region, and therefore, this method is spatially sensitive. The source-space projection allows calculating the grand averages of dipole moment time series across different subjects and measurement blocks thereby enhancing the signal-to-noise ratio canceling the uncorrelated system noise. The method is maximally sensitive for brain activity from sources at selected origins and orientations. Unwanted activities from more distant sources or sources having different orientations are combined less optimally, and therefore, the activity of these sources is reduced in the dipole moment waveforms.

In order to reduce the influence of interindividual neuroanatomical variations, the ASSR source locations in both conditions were referenced to the N1m source location. The distances of the estimated individual ASSR to the corresponding N1m source locations were calculated. The 95% confidence intervals were estimated for the 3-dimensional distance between ASSR and N1m sources across all subjects in both hemispheres in order to reject the null hypothesis of nondistinguishable distances between source locations. The null hypothesis was rejected in case if the mean value of the difference to N1m calculated for central beat ASSR (CB_ASSR) across all subjects was outside the 95% confidence ellipse of the mean distance to N1m calculated for the peripheral beat ASSR (PB_ASSR).

Furthermore, a Fourier analysis was performed on the single epoch source waveforms for each subject. Therefore, for each experimental condition, 6 times 256 stimulus-related epochs of 1-s length were concatenated to a single data vector. The Fast Fourier Transform applied to this data resulted in a high-resolution amplitude spectrum, which contained the evoked response as discrete lines with 1-Hz spacing and the noise distributed to 1536 lines within each 1-Hz frequency bin. The amplitude of the 40-Hz response was compared with 20 noise samples from the 1-Hz bins above and below 40 Hz, respectively. An F-test was used to test the null hypothesis that the 40-Hz amplitude was not different from the background noise (Lins et al. 1996; John and Picton 2000; Picton et al. 2003).

Results

ASSR Magnetic Field Distribution

Figure 2A illustrates the averaged ASSR field distribution data over 150 MEG channels recorded from a single subject. The responses to central beat are presented on the left side and to the peripheral beat on the right. The time interval of 100 ms contains 4 periods of the 40-Hz ASSR. As displayed in the figure, a dipolar ASSR structure is observed over each hemisphere. In the graphs of the Figure 2A, 2 maximally responding channels in the left and right hemispheres were plotted for comparison (note the different scaling left and right). For both conditions, the right hemisphere response was larger than that of the left hemisphere. A further observation was that the magnetic field amplitudes of the ASSR elicited by the peripheral beat were about 5 times larger than the ASSR to the central beat. The same time points after the stimulus onset have been chosen for the 2 types of stimuli and marked by arrows at the top of Figure 2. The isocontour maps of the field distributions corresponding to these time points are displayed in Figure 2B. The field distributions for the 2 different stimulus conditions were highly dipolar over both hemispheres, suggesting an explanation by single sources in each hemisphere. Also noticeable in this data representation is the dominance of the field strength above the right as compared with the left hemisphere. A common amplitude scale was used for the isocontour maps in order to illustrate the ASSR magnitude differences between both stimulus conditions.

Figure 2.

Auditory evoked magnetic fields in one subject. A: Waveforms of 150 MEG sensors for dichotic and diotic stimulation. The insets at the top show maximally responding sensors above the left and right hemispheres. B: Corresponding isocontour maps of the magnetic field distributions.

Figure 2.

Auditory evoked magnetic fields in one subject. A: Waveforms of 150 MEG sensors for dichotic and diotic stimulation. The insets at the top show maximally responding sensors above the left and right hemispheres. B: Corresponding isocontour maps of the magnetic field distributions.

Time Domain Analysis

The source-space projection method, applied to the averaged and 30- to 50-Hz band-pass–filtered magnetic field data resulted in grand averaged (n = 11) source waveforms shown in Figure 3 for the time interval from −0.2 to 1.2 s. The upper graph of the figure shows the stimulus waveform and the lower 2 panels show the grand averaged ASSR source waveforms from the right and the left hemispheres to peripheral and central beat, respectively. The PB_ASSR shows enhanced amplitude around 200 ms after the stimulus onset as compared with the almost constant amplitude between 0.4 and 1.0 s. The CB_ASSR generally follows similar time course although with considerably smaller amplitude. Group statistics indicate that the averaged right hemisphere source waveform was significantly larger than that of the left hemisphere (P ≤ 0.02 for the CB_ASSR and P ≤ 0.007 for the PB_ASSR).

Figure 3.

Grand averaged (n = 11) 30- to 50-Hz band-pass–filtered response waveforms of cortical source activity in the right hemisphere.

Figure 3.

Grand averaged (n = 11) 30- to 50-Hz band-pass–filtered response waveforms of cortical source activity in the right hemisphere.

Frequency Domain Analysis

Individual high-resolution Fourier spectra of cortical source activities under both stimulus conditions are illustrated in Figure 4 for the 30- to 50-Hz frequency range of interest. The spectral peaks at 40 Hz representing PB_ASSR and CB_ASSR were clearly above the background noise level (approximately 0.2 nAm). For both hemispheres, the PB_ASSR amplitudes were on the order of 2 nAm (upper panel of Fig. 4). Four side-lines with amplitudes above 0.5 nAm, which belong to the PB_ASSR, are distributed asymmetrically around the main 40-Hz peak with distances of multiples of 1 Hz and represent the amplitude fluctuations within the 1-s ASSR waveform. The amplitudes of all response components were slightly larger in the right as compared with the left hemisphere in the subject selected for illustration. The CB_ASSR spectrum (lower panel of Fig. 4) contains a single peak at 40 Hz with amplitude of 0.5 nAm in the right and smaller amplitude in the left hemisphere. Assuming that the noise is equally distributed across the frequencies, an F-test comparing the amplitude of the 40-Hz peak with the amplitudes of 40 sampled background noise lines around 40 Hz was applied. F-test ratios were calculated for each stimulus condition and each hemisphere. Results showed that the 40-Hz PB_ASSR peak was significant at P ≤ 0.0001 and F2,40 ≥ 501 in the right and at F2,40 ≥ 405 in the left hemisphere. The F values were considerably smaller for the CB_ASSR but still highly significant at P ≤ 0.0001 and F2,40 ≥ 12.3 for the right hemisphere and at P ≤ 0.0003 and F2,40 ≥ 8.7 for the left.

Figure 4.

Individual amplitude spectra of cortical source activity in the left and right hemispheres to peripheral and central beat stimulation.

Figure 4.

Individual amplitude spectra of cortical source activity in the left and right hemispheres to peripheral and central beat stimulation.

The polar plots in Figure 5 illustrate the individual ASSRs as vectors with the origin at the axes intersection. The length of the lines represents response amplitudes of each subject. The angle between a line and the x axis countering in clockwise direction corresponds to the difference between stimulus phase and response phase (Picton et al. 2003). The calculated circular variance according to Fisher (1993) suggests smaller intersubject variability for the PB_ASSR (circular variance 0.144 right and 0.137 left) than for the CB_ASSR (circular variance 0.334 right and 0.386 left). The complex amplitudes of CB_ASSR are spread in 3 quadrants and are with small amplitudes, whereas the complex amplitudes of PB_ASSR are spread in 2 quadrants and are with higher amplitudes. The result shows that ASSR to peripheral beat differ in amplitude and phase to ASSR to central beat.

Figure 5.

Individual complex ASSR amplitudes of cortical source activity evoked in left and right hemispheres by the peripheral and central beat stimuli. The group mean values are indicated with black squares.

Figure 5.

Individual complex ASSR amplitudes of cortical source activity evoked in left and right hemispheres by the peripheral and central beat stimuli. The group mean values are indicated with black squares.

Source Analysis

Magnetic source analysis of ASSR and the N1m transient responses to both stimulus conditions was carried out in all subjects. The source location of the N1m was used as an individual cortical marker and reference location for the estimation of the PB_ASSR and CB_ASSR source locations. To assess the significance of source distances, the mean across the subjects and the corresponding 95% confidence interval were calculated for each coordinate, as illustrated in Figure 6. The confidence ellipse for the ASSR did not include the N1m source and vice versa, indicating significant distances between ASSR and N1m source locations on the α = 0.05 level. This result was confirmed by the one-tailed Student t-test. Both PB_ASSR and CB_ASSR sources were more anterior, inferior, and medial compared with the N1m source. In the left hemisphere, the CB_ASSR and PB_ASSR were 0.5 and 1.0 cm more medial (P < 0.05) to the N1m source, respectively. Though the PB_ASSR and CB_ASSR sources were more inferior and anterior to the N1m, the differences did not reach statistical significance in the left hemisphere. In the right hemisphere, the PB_ASSR and CB_ASSR source location differences to the N1m source were more pronounced then in the left hemisphere, being about 5 mm more medial, more inferior, and more anterior. The distances between N1m and ASSR were significant in all directions of the coordinate system. The distance between CB_ASSR and PB_ASSR sources was largest in medial–lateral direction (about 5 mm) in the left hemisphere and with about the same amount in inferior–superior direction in the right hemisphere. The 95% confidence ellipses for the PB_ASSR do not include the CB_ASSR source and vice versa in the lateral–medial/anterior–posterior and lateral–medial/inferior–superior planes for the left hemisphere, indicating statistical significance for the distance between CB_ASSR and PB_ASSR sources (Fig. 6). In the right hemisphere, this is the case for the distance between the PB_ASSR and CB_ASSR sources in inferior–superior direction. The larger confidence ellipses in the left hemisphere illustrate the larger uncertainty of the source coordinates.

Figure 6.

Locations of the ASSR sources relative to the N1m sources in the Cartesian head coordinate system. The source locations in the left and right hemispheres are illustrated on the left and right graphs, respectively.

Figure 6.

Locations of the ASSR sources relative to the N1m sources in the Cartesian head coordinate system. The source locations in the left and right hemispheres are illustrated on the left and right graphs, respectively.

Discussion

This MEG study compared the ASSRs related to perception of beats of 2 tones interacting either in the auditory periphery or centrally. The sources of 40-Hz ASSR were located in the auditory cortices. Amplitudes of ASSR were larger for the peripheral beat than those for the central beat, and mean phases were different for both types of ASSR. A 2-tone complex shows envelope fluctuation at the difference frequency that resembles the envelope of an AM tone (Rutschmann and Rubinstein 1965). Such a stimulus produces beat perception and evokes a brain response following the beat frequency in the case of 2-tone stimulation or the modulation frequency in case of the AM tone (Picton et al. 1987; Dolphin and Mountain 1993; Henry 1996, 1998). The extraction of the envelope information, recorded as a brain response with a peak at the envelope frequency, was explained by Regan D and Regan MP (1988) as a demodulation process at the level of the auditory periphery. The acoustical signal is first rectified in the cochlea, due to the nonlinear function of the inner hair cells, and then low-pass filtered. Consequently, the responses from the brain stem and the auditory cortex have a spectral energy pattern that follows the envelope of the sound (Lins and Picton 1995). This explanation is mainly based on signal processing in linear systems. In our previous study (Ross et al. 2003), it was thought that the enhanced 40-Hz ASSR amplitudes to AM tones of low carrier frequencies and stronger interaction between multiple AM stimuli at low frequencies cannot be explained entirely with the characteristics of the auditory periphery and processing in a linear system. The frequency effects were different from those expected from the known frequency characteristics of middle-latency response and auditory brain stem response, as well as ASSR above 80 Hz. The involvement of 2 different mechanisms, a peripheral and a central one, has been suggested. Furthermore, the finding of ASSR reset induced by a concurrent stimulus (Ross et al. 2005a) or violation of the stimulus periodicity (Ross and Pantev 2004) supports the concept of a nonlinear system underlying the ASSR generation.

The current working hypothesis is that the ASSR represents oscillatory activity in a cortical network driven by periodic changes in the auditory input and responding maximally around 40 Hz. This concept is consistent with the observation that both AM and frequency-modulated tones evoke similar ASSR with similar generator origins being activated by separate auditory system parts (Picton et al. 1987). Therefore, we discuss the results of the current study in the context of 2 different stimulus attributes being processed by different parts of auditory system and yet activating 40-Hz oscillation (ASSR) in a common cortical network.

Place and Mechanism of ASSR Generation to a Central Beat

Recently, Schwarz and Taylor (2005) demonstrated in human EEG recordings cortical 40-Hz steady-state responses to dichotic stimulation with 2 tones around 400 Hz having an interaural difference of 40 Hz. They found that higher frequency tones around 3000 Hz were not efficient in generating the ASSR and that the ASSR corresponding to the 400 Hz dichotic stimulation was predominantly recorded from EEG electrodes of central and frontal brain areas. Based on these data, the authors assumed Heschl's gyri as a possible cortical projection of the ASSR to dichotic stimulation.

The dichotic stimulation in this experiment created 40-Hz oscillating interaural phase differences. The ability of the brain to detect this phase difference is manifested in the perception of a beat. After the tones have entered the cochlea, the dorsal and ventral cochlear nuclei obtain ipsilateral input only. All other structures along the auditory pathway receive bilateral inputs. Decussating in the auditory pathway begins at the superior olivary complex (SOC) (Levine 1981; Levine and Davis 1991) and continues through the lateral lemniscus and the inferior colliculus (IC) (Moore 1991). Neuronal activity with a period corresponding to the beat frequency has been recorded in cats by Wernick and Starr (1968). Their data suggested that the generation of the 40-Hz ASSR, corresponding to central beat, might start in the midbrain at the level of SOC. Using sinusoidal sounds around 600 Hz, beat frequencies between 20 and 63 Hz were generated, and therefore, the authors were able to record activity from SOC corresponding to several beat rates. This was the first evidence that response to central beat might appear already in SOC. Neurons distributed across the medial superior olive (MSO) and the lateral superior olive (LSO) were found to be sensitive to interaural phase and time differences (Fitzpatrick et al. 2002). According to the Jeffress' model (Jeffress 1948), such neurons are activated by an excitatory–excitatory (EE) binaural interaction. Therefore, these neurons can be regarded as coincidence detectors, discharging maximally when they receive simultaneous input from both sides (Batra and Yin 2004). Based on the principle of coincidence detection, Grothe (2000) investigated the spike rates of MSO neurons and found maximal rates when the inputs from left and right ears were in phase, generating phase-locked action potentials to the stimulus, and minimal when the bilateral inputs were out –of phase. Based on our experimental data, we propose the interaction of dichotically presented tones at the level of the MSO as a possible generation mechanism for the ASSR to central beats. If the difference between both tones is 40 Hz, each coincidence detector in the MSO detects 40 times per second equal phase for both inputs and 40 times out of phase inputs. Consequently, phase-sensitive MSO neurons respond periodically with a maximum spike rate every 25 ms.

It is known that sound localization based on detection of interaural time delay (ITD) in continuing sound relies on low frequencies only. Electrophysiological thresholds for interaural phase detection below 1500 Hz were recently demonstrated (Ross et al. 2007). An upper frequency limit on the order of 1500 Hz could explain why Schwarz and Taylor (2005) failed to show considerable ASSR using dichotic tones around 3000 Hz. However, additional work is necessary to confirm the assumed upper frequency limit for the generation of CB_ASSR close to 1500 Hz. Tollin and Yin (2005) suggested that EE and inhibitory–excitatory (IE) processing of time or phase delay in the LSO and MSO could interact additionally at the level of the IC, providing neurons with novel ITD sensitivity. Hence, in contrast to the Jeffress' model, based on EE binaural interaction occurring in the MSO, the processing of ITD also requires an IE mechanism in addition to EE binaural interaction involving the IC. Further, the activation of the sensitivity neurons to ITD is spread along higher stations of the auditory pathway to the fields of the auditory cortex as demonstrated by Reale and Brugge (1990). Hence, in addition to MSO as a possible place of generation of the response to a central beat, we also suggested involving of IC. We are aware that MEG recording of cortical ASSR is not able to identify the exact place of bilateral interaction. However, the MEG method described in this study allows investigating cortical correlates of binaural processing in the auditory brain stem and thus provides insight into auditory brain stem function in humans, otherwise inaccessible.

Sources of CB_ASSR and PB_ASSR were found to be located differently with a distance of about 5 mm in medial–lateral direction in the left hemisphere and with the same amount of distance in inferior–superior direction in the right hemisphere. Although the source distances were statistically significant on the α = 0.05 level, we would like to discuss this result carefully taking into account the limitations of the MEG method. Magnetic source modeling in MEG results in the estimation of a center of gravity of the underlying source distribution. We can determine this center location with high precision, but we do not know the actual source distribution in space. Even though, the centers of activity were different, we do not exclude the possibility that common neural network was involved in both CB_ASSR and PB_ASSR. This means, we interpret the experimental result of significantly different MEG source locations not in the way that 2 distinct neural populations exist, one responding to the peripheral beat and the other responding to the central beat. More likely, a common network is activated. However, our localization result indicates that the pattern of activation is different probably related to forming the different percepts of the 2 types of beat.

The location of ASSR sources relative to the N1 source location is consistent with previous studies and provides a good understanding of where the ASSR generating networks are located. The studies of Pantev et al. (1995) and Godey et al. (2001) reported that the N1 source was localized in the lateral parts of Heschl's gyrus and in the planum temporale (secondary auditory cortex). Hence, a shift of the ASSR source in anterior and medial direction, compared with N1m source, suggests that the ASSR is located rather in the primary auditory cortex (Engelien et al. 2000; Draganova et al. 2002; Ross et al. 2002). The finding that in both hemispheres the centrally and peripherally activated ASSR sources were significantly more medial, anterior, and inferior to the N1m source is a further argument that both types of ASSR are generated in primary auditory cortical areas (Pantev et al. 1993; Engelien et al. 2000).

Hemispheric Lateralization of ASSR

The left hemisphere is thought as being more specialized for language functions and the right for the processing of rhythmic sounds. As shown in this study, the 40-Hz ASSR amplitudes to both stimulus conditions were larger in the right hemisphere than in the left hemisphere, which is consistent with the above hypothesis and the fact that the right hemisphere is involved in spectral processing of sounds (Zatorre et al. 2002). It is also in line with the experimental results of Pardo et al. (1999), reporting that amplitude modulations in the middle of a pure tone generate stronger responses in the right than in the left hemisphere, as well as with the recent results of Ross et al. (2005b), who reported considerable right hemispheric laterality of ASSR to peripheral beats.

To summarize, this study investigated 2 different mechanisms of generation of 40 Hz ASSR to stimuli of 2 different perceptual categories, peripheral and central beat. Preliminary studies demonstrated extraction of a peripheral 40 Hz beat on the level of the cochlea spreading along the auditory pathway resulting in ASSR recorded from the cortex. Additionally, another central mechanism for generating of 40 Hz ASSR was assumed. The peripheral beat elicits the strong percept of a rough sound, whereas the central beat results in a very faint percept. The amplitude difference reflects the difference in strength of the percept. The mechanism of peripheral beat processing by the auditory system has been described in conjunction with signal processing in a linear system extracting the signal envelope first in the cochlea. This output signal activates maximally the cortex in agreement to the hypothesis of (Galambos et al. 1981). Producing a central beat as an alternative stimulus, we demonstrated evidence that 40 Hz cortical ASSR can result also from bilateral interaction. The central beat is generated as an output of neuronal impulses, phase-locked to the periodical time-delays in the stimulus. As it has been suggested, the neurons generating these impulses are specified for detection of time-delays and are most probably located in the LSO and MSO and their activation power is weaker. Comparison between ASSRs to peripheral and central beat showed that both types of beats activate common cortical network but they have different amplitude and phase, which might be due to the different auditory mechanism of generation.

Funding

Bundes Ministerium für Bildung und Forshung/Deutsches Zemtrum für Luft und Raumfahrt grant 01GW0520; Canadian Institute for Health Research; Canadian Foundation for Innovation.

The authors thank Patrick Bermudez for valuable comments on the manuscript. Conflict of Interest: None declared.

References

Batra
R
Yin
TC
Cross correlation by neurons of the medial superior olive: a reexamination
J Assoc Res Otolaryngology
 , 
2004
, vol. 
5
 (pg. 
238
-
252
)
Dolphin
WF
Mountain
DC
The envelope following response (EFR) in the Mongolian gerbil to sinusoidally amplitude-modulated signals in the presence of simultaneously gated pure tones
J Acoust Soc Am
 , 
1993
, vol. 
94
 (pg. 
3215
-
3226
)
Draganova
R
Ross
B
Borgmann
C
Pantev
C
Auditory cortical response patterns to multiple rhythms of AM sound
Ear Hear
 , 
2002
, vol. 
23
 (pg. 
254
-
265
)
Engelien
A
Schulz
M
Ross
B
Arolt
V
Pantev
C
A combined functional in vivo measure for primary and secondary auditory cortices
Hear Res
 , 
2000
, vol. 
148
 (pg. 
153
-
160
)
Fisher
NI
Statistical analysis of circular data
 , 
1993
Cambridge
Cambridge University Press
Fitzpatrick
DC
Kuwada
S
Batra
R
Transformations in processing interaural time differences between the superior olivary complex and inferior colliculus: beyond the Jeffress model
Hear Res
 , 
2002
, vol. 
168
 (pg. 
79
-
89
)
Galambos
R
Makeig
S
Talmachoff
PJ
A 40-Hz auditory potential recorded from the human scalp
Proc Natl Acad Sci USA
 , 
1981
, vol. 
78
 (pg. 
2643
-
2647
)
Godey
B
Schwartz
D
de Graaf
JB
Chauvel
P
Liegeois-Chauvel
C
Neuromagnetic source localization of auditory evoked fields and intracerebral evoked potentials: a comparison of data in the same patients
Clin Neurophysiol
 , 
2001
, vol. 
112
 (pg. 
1850
-
1859
)
Grothe
B
The evolution of temporal processing in the medial superior olive, an auditory brain stem structure
Prog Neurobiol
 , 
2000
, vol. 
61
 (pg. 
581
-
610
)
Henry
KR
Auditory nerve neurophonic produced by the frequency difference of two simultaneously presented tones
Hear Res
 , 
1996
, vol. 
99
 (pg. 
151
-
159
)
Henry
KR
Cochlear tuning in the gerbil: a comparison of responses to sinusoidal amplitude modulation and difference tone stimuli
Audiology
 , 
1998
, vol. 
37
 (pg. 
262
-
277
)
Jeffress
L
A place theory of sound localization
J Comp Physiol Psychol
 , 
1948
, vol. 
41
 (pg. 
35
-
39
)
John
MS
Lins
OG
Boucher
BL
Picton
TW
Multiple auditory steady-state responses (MASTER): stimulus and recording parameters
Audiology
 , 
1998
, vol. 
37
 (pg. 
59
-
82
)
John
MS
Picton
TW
MASTER: a Windows program for recording multiple auditory steady-state responses
Comput Methods Programs Biomed
 , 
2000
, vol. 
61
 (pg. 
125
-
150
)
Killion
MC
Insert earphones for more interaural attenuation
Hear Instrum
 , 
1985
, vol. 
36
 (pg. 
34
-
36
)
Levine
RA
Binaural interaction in brain stem potentials of human subjects
Ann Neurol
 , 
1981
, vol. 
9
 (pg. 
384
-
393
)
Levine
RA
Davis
PJ
Origin of the click-evoked binaural interaction potential, beta, of humans
Hear Res
 , 
1991
, vol. 
57
 (pg. 
121
-
128
)
Lins
OG
Picton
TW
Auditory steady-state responses to multiple simultaneous stimuli
Electroencephalogr Clin Neurophysiol
 , 
1995
, vol. 
96
 (pg. 
420
-
432
)
Lins
OG
Picton
TW
Boucher
BL
Durieux-Smith
A
Champagne
SC
Moran
LM
Perez-Abalo
MC
Martin
V
Savio
G
Frequency-specific audiometry using steady-state responses
Ear Hear
 , 
1996
, vol. 
17
 (pg. 
81
-
96
)
Moore
DR
Anatomy and physiology of binaural hearing
Audiology
 , 
1991
, vol. 
30
 (pg. 
125
-
134
)
Oster
G
Auditory beats in the brain
Sci Am
 , 
1973
, vol. 
229
 (pg. 
94
-
102
)
Pantev
C
Bertrand
O
Eulitz
C
Verkindt
C
Hampson
S
Schuierer
G
Elbert
T
Specific tonotopic organizations of different areas of the human auditory cortex revealed by simultaneous magnetic and electric recordings
Electroencephalogr Clin Neurophysiol
 , 
1995
, vol. 
94
 (pg. 
26
-
40
)
Pantev
C
Elbert
T
Makeig
S
Hampson
S
Eulitz
C
Hoke
M
Relationship of transient and steady-state auditory evoked fields
Electroencephalogr Clin Neurophysiol
 , 
1993
, vol. 
88
 (pg. 
389
-
396
)
Pardo
PJ
Makela
JP
Sams
M
Hemispheric differences in processing tone frequency and amplitude modulations
Neuroreport
 , 
1999
, vol. 
10
 (pg. 
3081
-
3086
)
Picton
TW
John
MS
Dimitrijevic
A
Purcell
D
Human auditory steady-state responses
Int J Audiol
 , 
2003
, vol. 
42
 (pg. 
177
-
219
)
Picton
TW
Skinner
CR
Champagne
SC
Kellett
AJ
Maiste
AC
Potentials evoked by the sinusoidal modulation of the amplitude or frequency of a tone
J Acoust Soc Am
 , 
1987
, vol. 
82
 (pg. 
165
-
178
)
Reale
RA
Brugge
JF
Auditory cortical neurons are sensitive to static and continuously changing interaural phase cues
J Neurophysiol
 , 
1990
, vol. 
64
 (pg. 
1247
-
1260
)
Regan
D
Regan
MP
The transducer characteristic of hair cells in the human ear: a possible objective measure
Brain Res
 , 
1988
, vol. 
438
 (pg. 
363
-
365
)
Ross
B
Draganova
R
Picton
TW
Pantev
C
Frequency specificity of 40-Hz auditory steady-state responses
Hear Res
 , 
2003
, vol. 
186
 (pg. 
57
-
68
)
Ross
B
Herdman
AT
Pantev
C
Stimulus induced desynchronization of human auditory 40-Hz steady-state responses
J Neurophysiol
 , 
2005
, vol. 
94
 (pg. 
4082
-
4093
)
Ross
B
Herdman
AT
Pantev
C
Right hemispheric laterality of human 40 Hz auditory steady-state responses
Cereb Cortex
 , 
2005
, vol. 
15
 (pg. 
2029
-
2039
)
Ross
B
Pantev
C
Auditory steady-state responses reveal amplitude modulation gap detection thresholds
J Acoust Soc Am
 , 
2004
, vol. 
115
 (pg. 
2193
-
2206
)
Ross
B
Picton
TW
Pantev
C
Temporal integration in the human auditory cortex as represented by the development of the steady-state magnetic field
Hear Res
 , 
2002
, vol. 
165
 (pg. 
68
-
84
)
Ross
B
Tremblay
KL
Picton
TW
Physiological detection of interaural phase differences
J Acoust Soc Am
 , 
2007
, vol. 
121
 (pg. 
1017
-
1027
)
Rutschmann
J
Rubinstein
L
Binaural beats and binaural amplitude-modulated tones: successive comparison of loudness fluctuations
J Acoust Soc Am
 , 
1965
, vol. 
38
 (pg. 
759
-
768
)
Schwarz
DW
Taylor
P
Human auditory steady state responses to binaural and monaural beats
Clin Neurophysiol
 , 
2005
, vol. 
116
 (pg. 
658
-
668
)
Tesche
CD
Uusitalo
MA
Ilmoniemi
RJ
Huotilainen
M
Kajola
M
Salonen
O
Signal-space projections of MEG data characterize both distributed and well-localized neuronal sources
Electroencephalogr Clin Neurophysiol
 , 
1995
, vol. 
95
 (pg. 
189
-
200
)
Tobias
JV
Application of a “Relative” procedure to a problem in binaural beat perception. Rep 63-17
Rep Civ Aeromed Res Inst US
 , 
1963
, vol. 
26
 (pg. 
1
-
8
)
Tollin
DJ
Yin
TC
Interaural phase and level difference sensitivity in low-frequency neurons in the lateral superior olive
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
10648
-
10657
)
Vrba
J
Robinson
SE
Signal processing in magnetoencephalography
Methods
 , 
2001
, vol. 
25
 (pg. 
249
-
271
)
Wernick
JS
Starr
A
Binaural interaction in the superior olivary complex of the cat: an analysis of field potentials evoked by binaural-beat stimuli
J Neurophysiol
 , 
1968
, vol. 
31
 (pg. 
428
-
441
)
Zatorre
RJ
Belin
P
Penhune
VB
Structure and function of auditory cortex: music and speech
Trends Cogn Sci
 , 
2002
, vol. 
6
 (pg. 
37
-
46
)