Tonotopic organization of the auditory cortex: pitch versus frequency representation

According to the place principles of the classical hearing theory, the physical entity frequency is encoded in the auditory periphery as place information (tonotopic representation), which is decoded in more central parts of the auditory system to form the subjective entity pitch. However, this relation is true only for pure-tone signals (spectral pitch); it can be quite different in the case of complex auditory stimuli (virtual pitch), thus requiring a multistage process for pitch formation. Neuromagnetic measurements showed that the tonotopic organization of the primary auditory cortex reflects the pitch rather than the frequency of the stimulus; that is, the pitch formation process must take place in subcortical regions.

Tinnitus remission objectified by neuromagnetic measurements

In a previous paper of ours (Hoke et al., 1989a) the hypothesis was put forward that the amplitude ratio of the two major waves of the auditory evoked magnetic field (AEF), M200/M100, is an objective measure which allows to discriminate between individuals suffering from tinnitus (ratio less than 0.5) and individuals without tinnitus (ratio greater than 0.5). We have now been able to trace the process of tinnitus remission in one exemplary case during a period of 256 days after acute onset of tinnitus (due to a noise trauma), in which the amplitude ratio recovered from 0 to a normal value of approximately 1. This very first objectification of tinnitus remission strongly supports our hypothesis and indicates that AEF may become an indispensable, invaluable tool in both tinnitus research and management.

Neuromagnetic evidence of an amplitopic organization of the human auditory cortex

It is well known that the location of the source of cortical auditory evoked responses, which can be determined neuromagnetically in humans using the concept of an equivalent current dipole (ECD), shifts with changing stimulus frequency ('tonotopic organization'). Not investigated so far, however, is the question of whether there exists also an 'amplitopic organization' of the human auditory cortex, i.e., a spatial distribution of neurons maximally responsive to respective 'best stimulus intensities.' We measured, in the study presented here, in 3 normally hearing subjects the auditory evoked magnetic field (AEF) in response to tone-burst stimulation with a carrier frequency of 1000 Hz at 6 different intensities (30-80 dB HL in 10 dB steps). The influence of stimulus intensity was quantified in terms of changes in the ECD parameters (amplitude, direction and spatial coordinates) which were determined such that a maximum correspondence between observed and calculated field distributions was obtained. The results of the neuromagnetic measurements presented here prove that the ECD location also shifts with changing stimulus intensity. The depth of wave M100 (latency of about 100 msec) decreases monotonically with increasing stimulus intensity while the horizontal ECD position is slightly shifted in the anterior direction. The results imply that, while topical mechanisms of frequency coding are similar at cortex and at the cochlear level, topical mechanisms of intensity coding are different at these levels.

Objective evidence of tinnitus in auditory evoked magnetic fields

The waveforms of the auditory evoked magnetic field in normal-hearing individuals and patients suffering from tinnitus are distinctly different. In tinnitus patients, the magnetic wave M200 (corresponding to the electric wave P200, or P2) is delayed and only poorly developed or even completely missing, while the amplitude of the magnetic wave M100 (corresponding to the electric wave N100, or N1) is significally augmented. A very characteristic feature turned out to be the amplitude ratio of the two waves M200 and M100. Below the age of 50, the amplitude ratio M200/M100 represents a clear-cut criterion to distinguish between tinnitus patients and individuals without tinnitus. In tinnitus patients, the ratio is less than 0.5, independent of age, whereas, in young and middle-aged normal-hearing individuals, it is greater than 0.5. Since in normal-hearing individuals the average amplitude ratio decreases linearly with age, the clusters of amplitude ratios of the two groups begin to overlap beyond the age of 50. The hypothesis is put forward that the decrease of the average amplitude ratio in normal-hearing individuals reflects a degenerative process, probably initiated by multiple exogenous and endogenous factors, which leads to sustained neural activity in the generators of wave M200 and eventually gives rise to the sensation of tinnitus. The absence or poor development of wave M200 is a concomitant phenomenon, resulting from the involved generators being less responsive to external stimuli.

Magnetic fields from the auditory cortex of a deaf human individual occurring spontaneously or evoked by stimulation through a cochlear prosthesis

In a postlingually deaf individual, the magnetic field evoked by stimulation through a cochlear prosthesis (extracochlear electrodes) as well as of the spontaneous magnetoencephalogram was measured over the hemisphere contralateral to the prosthesis (CP), and the results were compared with those obtained from normal-hearing subjects. The latency of the 2 best developed waves M100 and M200 turned out to be prolonged in the CP patient by approximately 40 ms. The amplitude of wave M100 was significantly diminished, while wave M200 was only poorly developed. Location and direction of the equivalent current dipole (ECD) calculated for wave M100 was in good agreement with normal data, whereas the dipole moment was only about one third of the average dipole moment found in normals. Furthermore, evidence was obtained for another magnetic field wave, preceding the delayed auditory wave M100, which exhibits the same latency, ECD location and direction as reported in the literature for the somatosensory evoked magnetic field. This wave probably results from stimulation, through the intratympanic electrodes, of somatosensory nerves innervating the tympanic cavity. A potential clinical application of neuromagnetic measurements is discussed: The calculation of the ECD moment from the auditory cortical magnetic field evoked by electrical stimulation at the promontory would allow to estimate, prior to CP implantation, the number of persisting, excitable nerve fibres.

Binaural interaction in brainstem auditory evoked potentials elicited by frequency-specific stimuli

The frequency specificity of the binaural interaction in brainstem auditory evoked potentials (BAEP) was investigated in ten normal-hearing young adults. A novel stimulus paradigm was devised to reduce the influence of the acoustic reflex (middle ear muscle contraction) on the BAEP, and to minimize the effect of variations in noise level. Sequences of six stimuli (rarefaction clicks or Gaussian-shaped tone pulses with carrier frequencies of 1, 2, 4 and 6 kHz) were periodically presented in the following order: right monaural, left monaural, binaural, left monaural, right monaural, binaural, with an interstimulus interval of 22 ms. Since the sequence of monaural stimuli with binaural stimuli interposed produces a uniform loudness and since the acoustic reflex is a consensual reflex, the relative high stimulus repetition rate (approx. 45/s) causes a muscle contraction which is equal on both sides and rather constant in time. This paradigm turned out to be usable for stimulus intensities as high as 80 dB nHL. The binaural difference potential (BDP) was computed by subtracting the sum of the monaurally (ipsilateral and contralateral) evoked potentials from the binaurally evoked potential. The major binaural interaction occurred in the latency range of BAEP waves V and VI, and there was no evidence of interaction in the earlier portion of the BAEP. Both latency and amplitude of the BDP components were evaluated statistically. The latency of the BDP components - except of the lasted one - showed an almost linear dependence both on stimulus intensity and stimulus frequency. The amplitude grew larger with decreasing frequency, and the visual detection threshold elevated as the stimulus frequency increased. Click stimuli, however, produced the largest amplitudes with lowest visual detection threshold. This novel stimulus paradigm appears to be most suitable for routine clinical investigations since high stimulus intensities can be used.

Tonotopic organization of the human auditory cortex revealed by transient auditory evoked magnetic fields

The tonotopic organization of the human auditory cortex has been investigated by systematic measurements of magnetic fields evoked by tone-bursts with carrier frequencies of 250, 500, 1000, 2000 and 4000 Hz. The measured field distribution changes with both time elapsed since stimulus onset and frequency of the stimulus. Nevertheless, the field distribution has always the same overall features and can be approximated by that of an equivalent current dipole located in a semi-infinite volume. This model can be described in terms of 5 parameter values: 3 orthogonal coordinates specifying the dipole location, and amplitude and angle of the dipole moment. The amplitude of the dipole moment is maximal at about 100 msec ('component 100m') and 160 msec ('component 160m') after stimulus onset. The depth estimated for the generator site of the 100m component shows a logarithmic dependence on test frequency whereas no similar behaviour could be observed for the 160m component. Anatomical studies performed in cadaver heads suggest that the equivalent current dipoles of both the 100m and the 160m component are located in the transverse temporal gyri.

Biomagnetic measurements using squids

Systematic studies of the magnetoencephalogram (MEG) in normal and pathological subjects (mainly with focal epilepsies) showed that the MEG may evidence significant brain activities even if they are not present in the electroencephalogram (EEG). They also showed that the MEG has a considerably higher spatial resolution than the EEG. A novel mapping technique was introduced to get such a representation of the data that would enable the investigator to draw his conclusions mainly from inspecting the plots. This technique is characterized by an isospectral amplitude (iso-SA) mapping of the scalp distribution of specified spectral components or frequency bands of the MEG power spectrum. With the above method we were able to localize an epileptiform focus using a noninvasive technique without applying an eliciting stimulus. Furthermore using SQUID measurements we were able to describe the behavior of the MEG when the brains of different subjects were subjected to low frequency sinusoidal binaural stimuli. Under these conditions it has been shown that the MEG tends to organize around discrete frequencies that depend on the interference pattern (beat) between the two inputs.

MEG measurements with SQUID as a diagnostic tool for epileptic patients

In experimental studies with a SQUID (Super-conducting QUantum Interference Device) second order gradiometer, we recently registered the magnetoencephalogram (MEG) from different subjects under different physiological and psychological conditions from which we will determine normal and abnormal function of the human brain. Thus with our first measurements using the MEG spectra, we have succeeded in identifying the exact location of the abnormality in the human brain as shown in several individuals.

Comparison between auditory brain stem responses evoked by rarefaction and condensation step functions and clicks

In order to evaluate the influence of the trailing edge of clicks on the auditory brain stem response (ABR) in normal ears, rarefaction and condensation step functions (RS and CS) compared to rarefaction and condensation clicks (RC and CC) at an intensity of 70 dB nHL were used. Significant intraindividual differences could be found for the latencies and amplitudes in the RS-CS, RS-RC and RC-CC comparison. However, the mean values of the complete group of test subjects showed no significant differences for the latencies and amplitudes, except the significantly greater amplitudes of wave I and II for R versus C step and R versus C click. Only a tendency to shorter latency for wave VI with R versus C step and click was revealed. These results show that there was no essential influence of the trailing edge of the used R and C clicks on the ABR. The latency of the ABR with excitation of the cochlea by step or click function seemed to be mainly determined by the internal oscillation sequence in the cochlea and not by the stimulus polarity.

Randomized data acquisition paradigm for the measurement of auditory evoked magnetic fields

The high variability of both amplitude and latency measures of the components of the auditory evoked magnetic field (AEMF), which we have attributed primarily to changes in the state of vigilance, makes it often impossible to compute significant isofield contour maps. Using a randomized data acquisition paradigm we have been able to considerably reduce the time-dependent fluctuations of the state of vigilance resulting in more stable latencies and in more stable and higher amplitudes of the AEMF components.

Comparison between simultaneously recorded auditory-evoked magnetic fields and potentials elicited by ipsilateral, contralateral and binaural tone burst stimulation

Both auditory-evoked magnetic fields (AEMF) and auditory-evoked potentials (AEP) mainly consist of three peaks with latencies of about 50, 100 and 160 ms. Comparison of responses to ipsilateral, contralateral and binaural stimulation yields no significant amplitude or latency differences of the AEP peaks whereas the simultaneously recorded AEMF peaks exhibit a 10 ms shorter latency and an approximately 38% greater amplitude for contralateral versus ipsilateral stimulation. This fact can be due to differences in the strength, location (especially the depth) and the direction of the dipole source, and a decision cannot be made considering the data recorded from just one position. Another finding is that binaural stimulation reduces the peak amplitudes by approximately 25% compared with contralateral stimulation. This result indicates some kind of interference between the ipsilateral and contralateral pathways ('binaural interaction').

Frequency-specific contributions to the auditory brain stem response derived by means of pure-tone masking

The pure-tone masking technique has been employed to determine the contributions of neural activity from different cochlear regions to the click-evoked auditory brain stem response (compound ABR). ABR to broad-band clicks were recorded without and with simultaneous presentation of pure tones of frequencies ranging from 8 to 0.5 kHz. Derived ABRs for individual frequency bands were obtained by subtracting the masked responses from the unmasked ones. To determine the contributions to the compound ABR, masked and derived ABRs were compared with unmasked responses. The frequency specificity of the contributions was more distinct at lower rather than at higher stimulus intensities. Independent of stimulus intensity, the 8- to 2-kHz regions turned out to be the main sources of waves I-V in the compound ABR, whereas waves VI and VII seem to be predominantly generated by contributions from regions specific for 1 and 0.5 kHz. The main advantage of the pure-tone masking technique as compared with the tone pip stimulation is that contributions from the low-frequency regions become more clearly detectable. The validity of this technique has been proved by comparing the compound ABR with the sum of the derived ABRs. The comparison of ABRs derived by pure-tone masking with those derived by conventional high-pass noise masking did prove the validity of the technique as well.

Comparison of the efficiency of various criteria for artifact rejection in the recording of auditory brain-stem responses (ABR)

Artifact rejection using various criteria has not yet been investigated comparatively in the determination of auditory brain-stem responses (ABR). The aim of this paper is to test the efficiency of three different criteria of artifact rejection: the 'level' criterion, the 'amplitude histogram' criterion, and the 'power' criterion, and to compare their practical applicability. With regard to the identification of the single waves of ABR (percentage frequency of their appearance) and their amplitudes without and with artifact rejection, it can be shown that the frequency of appearance increases severalfold when one of the artifact rejection criteria is used. The difference between amplitude means with artifact rejection (all three criteria) and those without artifact rejection is significant on the 5% level for waves I, II, and III and on the 1% level for wave V. When the three criteria are compared with regard to the amplitude means of the individual waves, it can be shown that the best results are obtained, when the 'level' criterion is used, then follows the 'power' criterion and third is the criterion of 'amplitude histogram'. The differences, however, are not significant. With reference to the 'level' criterion used, our investigation proved that the more complicated 'amplitude histogram' and 'power' criteria are not more efficient.

Derived brain stem responses by means of pure-tone masking

In order to demonstrate the possibility of deriving narrow-band responses, investigations using simultaneous pure-tone masking at frequencies of 1 and 0.5 kHz were carried out. Constant and reproducible derived responses were obtained at both masker frequencies within the following latency ranges: w6 (masker frequency 1 kHz) from 7.7 to 9.9 msec and w7 (masker frequency 0.5 kHz) from 9.2 to 11.2 msec at intensities of 70-20 dBSL. The agreement of our results with those obtained with the high-pass noise-masking technique permits us to ascribe our narrow-band responses (w6, w7) to the corresponding frequency regions around 1 and 0.5 kHz along the cochlear partition. This method is suitable for deriving frequency-specific responses without any sophistication.

Intramodal and interaural interactions of the human slow auditory evoked potential

It is proved that the human slow auditory evoked potential (SAEP) possess intramodal (frequency) and interaural specific properties. It is shown that tonotopic and bilateral organization of the SAEP generating system is better expressed at lower stimulus intensities. It is concluded that neuron populations of the SAEP generating system which perceive less different tone stimuli overlap each other greater than those which perceive more different stimuli. The data on intramodal and interaural specificity of SAEP are discussed from the point of view of its extralemniscal origin.