Reproducibility and Validity of Neuromagnetic Source Localization Using A Large Array Biomagnetometer

Reproducibility of measures of neurophysiological activity in Wernicke's area: a magnetic source imaging study

OBJECTIVE

The purpose of this study was to evaluate the reproducibility of estimates of neurophysiological activity obtained with Magnetic Source Imaging.

METHODS

Split-half data sets were obtained from 14 healthy volunteers during performance of a continuous recognition task for spoken words. The concurrent validity of spatiotemporal activation maps obtained with this task has been previously verified through comparisons with the Wada test and electrocortical stimulation mapping. Consecutive late activity sources (> 200 ms after stimulus onset) were modeled independently as equivalent current dipoles (ECDs) and used to identify the location of language-specific cortex in the left hemisphere (Wernicke's area).

RESULTS

Linear displacement of the geometric center of the cluster of ECDs in this region ranged between 2 and 8 mm across subjects. Intraparticipant variability (range) in the onset latency of activity was +/-50 ms, while the range of change in global field power for the entire set of ECDs in Wernicke's area was less than 17% in all cases.

CONCLUSIONS

The results indicate that despite its many conceptual limitations the ECD model can provide reliable estimates of regional cortical activity associated with the engagement of linguistic processes.

SIGNIFICANCE

The results highlight the need for reproducibility studies when research questions pose particular requirements for precision of estimates of regional neurophysiological activity.

An integrative MEG-fMRI study of the primary somatosensory cortex using cross-modal correspondence analysis

We develop a novel approach of cross-modal correspondence analysis (CMCA) to address whether brain activities observed in magnetoencephalography (MEG) and functional magnetic resonance imaging (fMRI) represent a common neuronal subpopulation, and if so, which frequency band obtained by MEG best fits the common brain areas. Fourteen adults were investigated by whole-head MEG using a single equivalent current dipole (ECD) and synthetic aperture magnetometry (SAM) approaches and by fMRI at 1.5 T using linear time-invariant modeling to generate statistical maps. The same somatosensory stimulus sequences consisting of tactile impulses to the right sided: digit 1, digit 4 and lower lip were used in both neuroimaging modalities. To evaluate the reproducibility of MEG and fMRI results, one subject was measured repeatedly. Despite different MEG dipole locations and locations of maximum activation in SAM and fMRI, CMCA revealed a common subpopulation of the primary somatosensory cortex, which displays a clear homuncular organization. MEG activity in the frequency range between 30 and 60 Hz, followed by the ranges of 20-30 and 60-100 Hz, explained best the defined subrepresentation given by both MEG and fMRI. These findings have important implications for improving and understanding of the biophysics underlying both neuroimaging techniques, and for determining the best strategy to combine MEG and fMRI data to study the spatiotemporal nature of brain activity.

Determination of activation areas in the human auditory cortex by means of synthetic aperture magnetometry

In this study we applied synthetic aperture magnetometry (SAM) to investigate active cortical areas associated with magnetically recorded transient and steady-state auditory evoked responses. For transient evoked responses, SAM images reveal an activated volume of cortical tissue within the lateral aspect of the superior temporal plane. The volume of cortical activation for steady-state responses was located more medially than that for transient evoked responses. Additionally, SAM also reveals a small overlap of activated areas between transient and steady-state evoked responses, which has not be demonstrated when using equivalent current dipole (ECD) source modeling. Source waveforms from SAM and ECD analyses show comparable temporal information. Results from this study suggest that SAM is a useful technique for imaging cortical structures involved in processing perceptual information.

Studies of tonotopy based on wave N100 of the auditory evoked field are problematic

There is still dissension as to whether the auditory evoked field (AEF) reflects tonotopy in the auditory cortex. That notwithstanding, particularly the pronounced AEF wave occurring about 100 ms after stimulus onset (N100 m) is increasingly used for the investigation of issues such as cortical reorganization and representation of virtual pitch. Thus, it appears to be time for a critical revaluation of the supposed tonotopic organization of the N100 m generator. In the present magnetoencephalography study, the response to tonebursts of 500 ms duration, monaurally presented 60 dB above threshold, was recorded with a 37-channel axial gradiometer system over the hemisphere contralateral to the side of stimulation. The stimulus frequencies were 250, 500, 1000, and 2000 Hz. About 250 stimuli of each type were presented in random order in four independent sessions at intervals uniformly distributed between 2 and 2.8 s. An analysis of 19 hemispheres in 10 normal-hearing subjects showed a high intraindividual reproducibility, but also a substantial interindividual variability. In most cases, the dipole location either exhibited no significant frequency dependence at all, the dipoles for the four frequencies were not orderly aligned, or the data disagreed with the single-dipole model. In the few cases showing an arrangement of dipoles consistent with the assumption of an orderly tonotopic cortical map, the most relevant coordinate varied from subject to subject. Regarding theses results, it seems crucial to understand wave N100 m on the basis of individual subjects, whereas conclusions relying on mean dipole locations for groups of subjects are problematic.

Effects of sleep on pain-related somatosensory evoked magnetic fields in humans

We investigated the effects of sleep on pain-related somatosensory evoked magnetic fields (SEFs) following painful electrical stimulation to identify the mechanisms generating them in both fast A-beta fibers relating to touch and slow A-delta fibers relating to pain. While the subjects were awake, non-painful and painful electrical stimulations were applied, and while asleep, painful stimulation was applied to the left index finger. During awake, five components (1M-5M) were identified following both non-painful and painful stimulation, but the 4M and 5M at around 70-100 ms and 140-180 ms, respectively, were significantly enhanced following painful stimulation. During sleep, 1M and 2M generated in the primary somatosensory cortex (SI) did not show a significant change, 3M in SI showed a slight but significant amplitude reduction, and 4M and 5M generated in both SI and the secondary somatosensory cortex (SII) were significantly decreased in amplitude or disappeared. The 4M and 5M are complicated components generated in SI and SII ascending through both A-beta fibers and A-delta fibers. They are specifically enhanced by painful stimulation due to an increase of signals ascending through A-delta fibers, and are markedly decreased during sleep, because they much involve cognitive function.

Localization accuracy of single current dipoles from tangential components of auditory evoked fields

We investigated the localization accuracy of single current dipoles from the tangential components of auditory evoked fields. The tangential fields were measured using planar gradiometers arranged in a way so as to detect two orthogonal field components parallel to a flat plane. Field responses to 1 kHz pure tones were recorded and equivalent current dipoles (ECDs) of N1m peak were estimated based on a locally fitted spherical conductor model. As a measure of localization accuracy, the standard deviation of the coordinates of the ECDs of N1m was obtained from repeated measurements for one subject. The estimated ECDs had a standard deviation of 5.5 mm and their mean location was at the supratemporal plane in the sylvian fissure on the MR image of the subject. In order to investigate the contribution of various errors to the localization accuracy, simulations using a sphere model and experiments using a realistically shaped skull phantom were performed. It was found that the background noise, which consisted of instrumental noise and spontaneous brain fields, was the main source of the errors that could explain the observed standard deviation. Further, the amount of systematic error was much less than the standard deviation due to the background noise. These results suggest that the volume currents in a non-spherical conductor shape such as the temporal region do not produce substantial errors in the localization of current dipoles from tangential components of auditory evoked fields.

Visual detection of motion speed in humans: spatiotemporal analysis by fMRI and MEG

Humans take a long time to respond to the slow visual motion of an object. It is not known what neural mechanism causes this delay. We measured magnetoencephalographic neural responses to light spot motion onset within a wide speed range (0.4-500 degrees /sec) and compared these with human reaction times (RTs). The mean response latency was inversely related to the speed of motion up to 100 degrees /sec, whereas the amplitude increased with the speed. The response property at the speed of 500 degrees /sec was different from that at the other speeds. The speed-related latency change was observed when the motion duration was 10 msec or longer in the speed range between 5 and 500 degrees /sec, indicating that the response is directly related to the speed itself. The source of the response was estimated to be around the human MT+ and was validated by functional magnetic imaging study using the same stimuli. The results indicate that the speed of motion is encoded in the neural activity of MT+ and that it can be detected within 10 msec of motion observation. RT to the same motion onset was also inversely related to the speed of motion but the delay could not be explained by the magnetic response latency change. Instead, the reciprocal of RT was linearly related to the reciprocal of the magnetic response latency, suggesting that the visual process interacts with other neural processes for decision and motor preparation.

Neuromagnetic source localization of auditory evoked fields and intracerebral evoked potentials: a comparison of data in the same patients

OBJECTIVE

To compare the localizations of different neural sources (a) obtained from intracerebral evoked responses and (b) calculated from surface auditory evoked field responses recorded in the same subjects. Our aim was to evaluate the resolving power of a source localization method currently used in our laboratory, which is based on a recent spatio-temporal algorithm used in magneto-encephalography (MEG).

METHODS

Auditory evoked responses were studied in 4 patients with medically intractable epilepsy. These responses were recorded from depth electrodes implanted in the auditory cortex for pre-surgical evaluation (stereo-electro-encephalography (SEEG)), as well as from surface captors (for MEG) placed on the scalp after removal of the depth electrodes. Auditory stimuli were clicks and short tone bursts with different frequencies.

RESULTS

All middle-latency components (from 13 to 70 ms post-stimulus onset) were recorded and localized (via SEEG) along Heschl's gyrus (HG). MEG reliably localized Pam and P1m in the same area of HG that intracerebral recordings localized them in. No significant delay between SEEG and MEG latencies was observed. Both methods suggest that N1 is generated from different sources in the intermediate and lateral parts of the HG and in the planum temporale (PT). The source of P2 (PT and/or Area 22) remains unclear and was in one case, localized in different regions according to the method used. This latter component may therefore also be generated by different sources.

CONCLUSIONS

The results suggest that both techniques are useful and may be used together in a complementary fashion. Intracerebral recordings allow the researcher to validate and interpret surface recordings.

The somatosensory evoked magnetic fields

Averaged magnetoencephalography (MEG) following somatosensory stimulation, somatosensory evoked magnetic field(s) (SEF), in humans are reviewed. The equivalent current dipole(s) (ECD) of the primary and the following middle-latency components of SEF following electrical stimulation within 80-100 ms are estimated in area 3b of the primary somatosensory cortex (SI), the posterior bank of the central sulcus, in the hemisphere contralateral to the stimulated site. Their sites are generally compatible with the homunculus which was reported by Penfield using direct cortical stimulation during surgery. SEF to passive finger movement is generated in area 3a or 2 of SI, unlike with electrical stimulation. Long-latency components with peaks of approximately 80-120 ms are recorded in the bilateral hemispheres and their ECD are estimated in the secondary somatosensory cortex (SII) in the bilateral hemispheres. We also summarized (1) the gating effects on SEF by interference tactile stimulation or movement applied to the stimulus site, (2) clinical applications of SEF in the fields of neurosurgery and neurology and (3) cortical plasticity (reorganization) of the SI. SEF specific to painful stimulation is also recorded following painful stimulation by CO(2) laser beam. Pain-specific components are recorded over 150 ms after the stimulus and their ECD are estimated in the bilateral SII and the limbic system. We introduced a newly-developed multi (12)-channel gradiometer system with the smallest and highest quality superconducting quantum interference device (micro-SQUID) available to non-invasively detect the magnetic fields of a human peripheral nerve. Clear nerve action fields (NAFs) were consistently recorded from all subjects.

Magnetoencephalographic studies of functional organization and plasticity of the human auditory cortex

Magnetoencephalography has proven to be a powerful noninvasive tool for investigating the functional organization of the human auditory cortex and its plastic changes. The first part of this review summarizes some recent experiments on the tonotopic organization, which can be observed not only in the slow auditory evoked fields, but also in the middle-latency and the steady-state fields. In the second part of this review, recent studies on plasticity of the auditory cortex are outlined. These studies showed that the cortical representation of tones may change within hours after a reversible "functional deafferentation" (short-term plasticity) and that early musical training leads to an expansion in the cortical representation of complex harmonic sounds (long-term plasticity).

Dermatome versus homunculus; detailed topography of the primary somatosensory cortex following trunk stimulation

OBJECTIVE

Identification of a detailed topography of the receptive area for each of the thoracic dermatomes in humans using somatosensory evoked magnetic fields (SEF).

METHODS

We analyzed the location of the equivalent current dipole (ECD) of SEF following electrical stimulation of the skin at Th4, Th6, Th8, Th10 and Th12 dermatomes in 14 normal subjects.

RESULTS

Three deflections, M18, M25 and M40, were obtained within 60 ms of stimulation of Th6, Th8 and Th10 dermatomes. No consistent deflection could be identified following Th4 and Th12 dermatomal stimulation, probably due to a poor signal-to-noise ratio and difficulty in fixing the stimulation electrodes. M18 was absent or small in amplitude. The latency of M25 ranged from short to long in the order Th6, Th8 and Th10 (P<0.05). ECDs of all components for each site stimulation were located in the truncal area of the primary somatosensory cortex. Although the locations of the ECDs tend to be arranged from lateral to medial in the sequence Th6, Th8 and Th10, the difference was not significant.

CONCLUSION

The representation area of the trunk is small, and the receptive areas for the stimulation of Th6, Th8 and Th10 dermatomes are considered to be very close or to overlap.

Magnetoencephalographic localization in pediatric epilepsy surgery: comparison with invasive intracranial electroencephalography

The object of this study was to determine the concordance of the anatomical location of interictal magnetoencephalographic (MEG) spike foci with the location of ictal onset zones identified by invasive ictal intracranial electroencephalographic recordings in children undergoing evaluation for epilepsy surgery. MEG was performed in 11 children with intractable, nonlesional, extratemporal, localization-related epilepsy. Subsequently, chronic invasive intracranial electroencephalographic monitoring was performed by using subdural electrodes to localize the ictal onset zone and eloquent cortex. Based on the invasive monitoring data, all children had excision of, or multiple subpial transections through, ictal onset cortex and surrounding irritative zones. In 10 of 11 patients, the anatomical location of the epileptiform discharges as determined by MEG corresponded to the ictal onset zone established by ictal intracranial recordings. In all children, the anatomical location of the somatosensory hand area, determined by functional mapping through the subdural electrode array, was the same as that delineated by MEG. Nine of 11 patients became either seizure-free or had a greater than 90% reduction in seizures after surgery, with a mean follow-up of 24 months. MEG is a powerful and accurate tool in the presurgical evaluation of children with refractory nonlesional extratemporal epilepsy.

Short-term plasticity of the human auditory cortex

Magnetoencephalographic measurements (MEG) were used to examine the effect on the human auditory cortex of removing specific frequencies from the acoustic environment. Subjects listened for 3 h on three consecutive days to music "notched" by removal of a narrow frequency band centered on 1 kHz. Immediately after listening to the notched music, the neural representation for a 1-kHz test stimulus centered on the notch was found to be significantly diminished compared to the neural representation for a 0.5-kHz control stimulus centered one octave below the region of notching. The diminished neural representation for 1 kHz reversed to baseline between the successive listening sessions. These results suggest that rapid changes can occur in the tuning of neurons in the adult human auditory cortex following manipulation of the acoustic environment. A dynamic form of neural plasticity may underlie the phenomenon observed here.

Modification of auditory pathway functions in patients with hearing improvement after middle ear surgery

We recorded auditory-evoked magnetic responses with a whole-scalp 122-channel neuromagnetometer from seven adult patients with unilateral conductive hearing loss before and after middle ear surgery. The stimuli were 50-msec 1-kHz tone bursts, delivered to the healthy, nonoperated ear at interstimulus intervals of 1, 2, and 4 seconds. The mean preoperative pure-tone average in the affected ear was 57 dB hearing level; the mean postoperative pure-tone average was 17 dB. The 100-msec auditory-evoked response originating in the auditory cortex peaked, on average, 7 msecs earlier after than before surgery over the hemisphere contralateral to the stimulated ear and 2 msecs earlier over the ipsilateral hemisphere. The contralateral response strengths increased by 5% after surgery; ipsilateral strengths increased by 11%. The variation of the response latency and amplitude in the patients who underwent surgery was similar to that of seven control subjects. The postoperative source locations did not differ noticeably from preoperative ones. These findings suggest that temporary unilateral conductive hearing loss in adult patients modifies the function of the auditory neural pathway.

Timing of motion representation in the human visual system

Visual stimulus in apparent motion evokes a magnetic field from the extrastriate cortex in humans. To investigate what this magnetic field represents, we measured the latencies of the responses in three subjects to the stimuli in apparent motion at various spatial separations. These different latencies were inversely related to the spatial separations of the stimuli (range of 74 to 182 ms) and correlated with each subject's reaction time. The direction of motion affected neither the latency of the magnetic response nor the reaction times. Estimations of the origins of the evoked magnetic fields showed they were always in the same area. In two subjects, the sites were around the meeting point of the ascending limb of the inferior temporal sulcus and the lateral occipital sulcus. In the third subject, the site was in the vicinity of the angular gyrus. The difference between the magnetic response and reaction time was fairly constant (about 64 ms) among the subjects. We consider the magnetic response to be related to the generation of a motion image: First, the response clearly corresponded to human reaction times to the same stimuli: Second, the fact that the magnetic response was related to the spatial separations but independent of the direction of motion is not explained if the response is evoked simply by both the onset and offset of the object in the stimulus. Furthermore, individual reaction times were mainly delayed by the speed of the process that generated the motion image.

Replicability of MEG and EEG measures of the auditory N1/N1m-response

We investigated the replicability of the source location, amplitude and latency measures of the auditory evoked N1 (EEG) and N1m (MEG) responses. Each of the 5 subjects was measured 6 times in two recording sessions. Responses to monaural stimuli were recorded from 122 MEG and 64 EEG channels simultaneously. The EEG data were modeled with a symmetrically-located dipole pair. For the MEG data, one dipole in each hemisphere was located independently using a subset of channels. Standard deviation (SD) was used as a measure for replicability. The average SD of the x, y and z coordinates of the contralateral N1m dipole was about 2 mm, whereas the corresponding figures for the ipsilateral N1m and the contra- and ipsilateral N1 were about twice as large. The SDs of the dipole amplitudes and latencies were almost equal with MEG and EEG. The amplitude and latency measures of the MEG field gradient waveforms were almost as replicable as those of the dipole models. The results suggest that both MEG and EEG can be used for investigating the simultaneous activity of the left and right auditory cortices independently, MEG being superior in certain experimental setups.

Dipole source localization by means of maximum likelihood estimation. II. Experimental evaluation

A dipole source analysis of repeated measurements of the auditory evoked field (AEF) elicited by 60 dB SL tonebursts confirmed the previous theoretical prediction that a maximum likelihood estimation technique reduces localization errors resulting from noise in the measured data by about a factor of two as compared to the common least-squares fit procedure. To achieve a comparable improvement by averaging independent epochs (Gaussian noise assumed), the number of averages would have to be increased by a factor of 4. Conversely, the measurement time required to achieve a given localization accuracy could be reduced by that factor.

Reproducibility and validity of electric source localisation with high-resolution electroencephalography

The present study investigates the reproducibility and validity of the EEG source localisation of somatosensory evoked potentials (SEPs) using high-resolution EEG (61 scalp electrodes) and a source reconstruction on the basis of the individual brain morphology as obtained from magnetic resonance images (MRIs). The somatosensory evoked potentials (SEPs) to electrical stimulation of the right median nerve were repeatedly collected from the scalp of one healthy subject in 9 replications run on 9 different days. The source reconstruction for the 19 ms SEP component was performed by using a single moving dipole model as a source model. Two different head models were used: a spherical 3 shell model and a more realistically shaped 3 compartment model computed using the boundary element method (BEM). The source locations of the 19 ms SEP component were found to be highly reproducible using both head models: the mean standard deviation of the dipole locations was found to be 2.6 mm for the 3 shell model and 4 mm for the more realistically shaped head model. By projection into the individual MRI, the dipoles resulting from either head models were found to be located within the postcentral gyrus. The electric source locations were consistent with the maximum of the task-specific changes seen in a functional magnetic resonance imaging (fMRI) experiment when using the same somatosensory stimulation protocol.

Magnetoencephalography may help to improve functional MRI brain mapping

The validity of functional magnetic resonance imaging (FMRI) brain maps with respect to the sites of neuronal activation is still unknown. One source of localization error may be pixels with large signal amplitudes, since such pixels may be expected to overlie large vessels, running remote from the centre of neuronal activation. In this study, magnetoencephalography was used to determine the centre of neuronal activation in a simple finger tapping task. The localization accuracy of conventional FMRI depending on FMRI signal enhancement was investigated relative to the magnetoencephalography reference. The results show a deterioration of FMRI localization with increasing signal amplitude related to increased contributions from large vessels. We conclude that FMRI data analysis should exclude large signal amplitudes and that magnetoencephalography may help to improve FMRI brain mapping results in a multimethod approach.

Possibilities and limitations of magnetic source imaging of methohexital-induced epileptiform patterns in temporal lobe epilepsy patients

The usefulness of MEG-based techniques in lateralizing and localizing the epileptogenic area was investigated in the present study. Spontaneous and methohexital-induced spikes were studied in a group of 15 patients with temporomesial epilepsy using a 37-channel neuromagnetometer. The accuracy of the magnetic source imaging was compared to the results of electrocorticographic (ECoG) recordings. Differences of drug-induced spike densities in the MEG recordings between both sides confirmed a similar lateralizing power of the MEG and ECoG recordings. Source location analyses based on a moving dipole model resp. a rotating dipole model were performed using a spherical head model. After subdivision of the volume of each patient's head, 8 cm3 cubicles containing at least 3 source locations were projected onto the individual MRI scan and resulted in source locations within or close to the presurgically defined primary epileptogenic area only in 3 of the 15 patients. Spike induction by methohexital has the advantage of shortening the recording period as compared to recordings of interictal epileptiform discharges. However, the correlation analyses of spike densities from MEG and ECoG recordings and the source location analyses from MEG recordings indicate that spike generated in deep temporomesial structures may escape the MEG registration.

Mismatch field to tone pairs: neuromagnetic evidence for temporal integration at the sensory level

The mismatch field (MMF) to minor pitch changes in two experimental conditions was studied. Standard tones of 1000 Hz and deviant tones of 1050 Hz both of 50 ms duration were delivered in single tone condition. Paired tones of the same duration were used in the paired tone condition. The standard tone pair consisted of two 1000 Hz tones, whereas the deviant tone pair was composed of a 1000 Hz tone in the first position and a 1050 Hz tone in the second position with a silent interval of 15 ms between the two. Standards of 90% and deviants of 10% probability were presented in random order and with a randomized interstimulus interval between 600 and 900 ms. The source analysis showed a more lateral location for the MMF obtained in the paired tone condition (MMF.P) compared to the MMF elicited by the single deviants (MMF.S). The source location of both the MMF.P and MMF.S turned out to be significantly anterior relative to the sources of the M100. The increased stimulus repetition in the paired tone condition (two times more stimuli than in the single tone condition) lead to a strong suppression of the field amplitude and of the dipole moment of the M100, while this effect could not be seen for the MMF. The data demonstrate a fundamental difference between the processes reflected by the M100 and the MMF: while the M100 represents the processing of every individual tone, the MMF reflects the change detection of the paired stimuli as unitary events, forming a perceptual group. The different sources of the MMF.P and MMF.S also support an integrated processing of the paired stimuli.

Tonotopic organization of the sources of human auditory steady-state responses

Steady-state responses (SSRs) or steady-state fields (SSFs) show maximum amplitude when tone pulses are presented at repetition rates near 40 Hz. This result has led to the hypothesis that the SSR/SSF consists of superimposed transient 'middle latency' responses which display wave periods near 40 Hz and summate with one another when phase locked by 40 Hz steady-state stimulation. We evaluated this hypothesis by comparing the cortical sources of the 40 Hz auditory SSF with sources of the middle latency Pa wave which is prominent in electrical and magnetic recordings, and with the cortical sources of the familiar N1 wave, at different carrier frequencies between 250 and 4000 Hz. SSF sources determined for the different carrier frequencies were found to display a 'medial' tendency tonotopy resembling that of the N1m (sources for the higher frequencies represented more deeply within the supratemporal sulcus), opposite the 'lateral' tendency tonotopy of the middle latency Pam (sources for the higher frequencies situated more laterally). A medial SSF tonotopy was observed in each of the subjects investigated, including three subjects for whom Pam and N1m maps were also available. These findings suggest that the 40 Hz SSF may not consist of summated or entrained middle latency responses, as has previously been proposed. Alternative mechanisms for the SSR are discussed.

Effects of sleep on somatosensory evoked responses in human: a magnetoencephalographic study

We studied the effects of sleep on somatosensory evoked magnetic fields (SEFs) following median nerve stimulation in normal subjects, to investigate the changes of functional processing of sensory perception in the primary and second sensory cortices (SI and SII). The early components, 1M, 2M and 3M, which were generated in SI contralateral to the stimulated nerve, showed no significant change of latency or amplitude in stage 1 or 2 as compared with those in the awake state. The long-latency response, 4M whose latency was about 100 ms, was significantly enhanced in stage 2. The 4M was considered to be generated in SI and SII in the awake state, but the enhanced 4M in stage 2 was restricted in SI. The 4M(I) generated in SII of the hemisphere ipsilateral to the stimulated nerve, corresponding to 4M in the contralateral hemisphere, was absent during sleep. These findings were probably due to the difference of activities between SI and SII during sleep, that is, an increase of sensitivity to somatosensory stimulation in SI but a decrease or disappearance in SII.

Current dipole localization with an ideal magnetometer system

The goal of the study was to explore the most fundamental aspects of a magnetoencephalography (MEG)-based dipole source analysis. For that purpose, a MEG measurement with an ideal magnetometer system (providing the radial component of the magnetic field as a continuous function) is considered. The analytical formulas derived for the variances and covariances of the parameter estimation errors, validated by means of Monte Carlo simulations, allow quantitative predictions in terms of dipole depth, radius and span of the magnetometer system, signal-to-noise (SNR) ratio, and other parameters. A negative correlation exists between radial coordinate and longitudinal component of the moment (perpendicular to radial direction, same plane as actual dipole moment and center of sphere), whereas the other parameters are independent. The standard deviations of the five dipole parameters show fundamental differences with respect to their asymptotic behavior for deep dipoles: If the root mean square (rms) value of the magnetic field is kept constant (moment with depth-dependent amplitude), the error for the transverse coordinate (perpendicular to radial and longitudinal coordinate) is proportional to the distance R between dipole and center of sphere, the errors for the other dipole coordinates, and the relative error for the transverse component of the dipole moment are constant, and the relative error for the longitudinal component of the dipole moment follows a 1/R law.

Binaural fusion and the representation of virtual pitch in the human auditory cortex

The auditory system derives the pitch of complex tones from the tone's harmonics. Research in psychoacoustics predicted that binaural fusion was an important feature of pitch processing. Based on neuromagnetic human data, the first neurophysiological confirmation of binaural fusion in hearing is presented. The centre of activation within the cortical tonotopic map corresponds to the location of the perceived pitch and not to the locations that are activated when the single frequency constituents are presented. This is also true when the different harmonics of a complex tone are presented dichotically. We conclude that the pitch processor includes binaural fusion to determine the particular pitch location which is activated in the auditory cortex.

Internal consistency of dipole localizations for the human movement-evoked magnetic field component 1 (MEF 1)

The present magnetoencephalographic study was conducted in order to assess the accuracy of dipole localizations for the movement-evoked field component 1 (MEF 1). Three male subjects were requested to perform self-paced flexions of their index finger and thumb in repeated sessions of 60 trials while the neuromagnetic field was recorded by a 31 channel system. Single moving dipole localizations were performed for the MEF 1. The error within single sessions was calculated by split-half reliability and window-homogeneity in a total of 61 sessions. The mean spatial deviation between both halves amounted to 3.8 mm. The window-homogeneity was found to be 2 mm deviation/10 ms.

Separation of cardiac conduction system and atrial activities by spatial regression

Multiple channel magnetocardiography is potentially useful for the study of the cardiac conduction system. However, normal atrial repolarization occurs simultaneously and obscures the interpretation of the net signal. Magnetocardiographic data in four normal subjects at rest and mild exercise were found to exhibit high spatial correlation during atrial activation. Based on measured channel-to-channel covariances, the atrial repolarization signals as measured in channels in the null zone of conduction system activity were used to estimate atrial repolarization in all channels. A linear prediction method was used which based on the "kriging" estimator of geostatistical theory. Due to high spatial correlations in the limited thoracic region studied, predictions based on a single null channel were found to be adequate. Removal of the atrial component facilitates the beat-by-beat estimation of conduction system changes during rest and exercise using the multiple channel biomagnetometer.

Potential contribution of bilateral magnetic source imaging to the evaluation of epilepsy surgery candidates

The current procedures that are used to evaluate candidates for epilepsy surgery are time-consuming, costly, and often invasive. Magnetic source imaging (MSI), the combination of magnetoencephalography and anatomic imaging modalities, has shown promise as an efficient noninvasive means of localizing and characterizing seizure sources for possible resection. However, MSI has been limited by the inability to conduct simultaneous bilateral monitoring. In this study, a newly developed dual-magnetometer system was employed to record bilaterally the interictal activity in 30 candidates for epilepsy surgery. A standard monitoring protocol that included concurrent electroencephalographic recording and required a 2- to 3-hour examination period for each patient was developed. As a first step in a series of studies, the resultant MSI indications were compared with the information available from standard magnetic resonance imaging and concurrent electroencephalographic results. In 83% of the cases, this MSI protocol provided new information about the location of interictal epileptic activity that could be directive for subsequent patient care. Based on these results, it seems that MSI may become a cost-effective early step in epilepsy surgery evaluation. To continue the development on this basis, a study intended to validate the accuracy of MSI indicated by comparison with invasive electroencephalography has been initiated.

Specific tonotopic organizations of different areas of the human auditory cortex revealed by simultaneous magnetic and electric recordings

This paper presents data concerning auditory evoked responses in the middle latency range (wave Pam/Pa) and slow latency range (wave N1m/N1) recorded from 12 subjects. It is the first group study to report multi-channel data of both MEG and EEG recordings from the human auditory cortex. The experimental procedure involved potential and current density topographical brain mapping as well as magnetic and electric source analysis. Responses were compared for the following 3 stimulus frequencies: 500, 1000 and 4000 Hz. It was found that two areas of the auditory cortex showed mirrored tonotopic organization; one area, the source of N1m/N1 wave, exhibited higher frequencies at progressively deeper locations, while the second area, the source of the Pam/Pa wave, exhibited higher frequencies at progressively more superficial locations. The Pa tonotopic map was located in the primary auditory cortex anterior to the N1m/N1 mirror map. It is likely that N1m/N1 results from activation of secondary auditory areas. The location of the Pa map in A1, and its N1 mirror image in secondary auditory areas is in agreement with observations from animal studies.

Detection of stimulus deviance within primate primary auditory cortex: intracortical mechanisms of mismatch negativity (MMN) generation

Mismatch negativity (MMN) is a cognitive, auditory event-related potential (AEP) that reflects preattentive detection of stimulus deviance and indexes the operation of the auditory sensory ('echoic') memory system. MMN is elicited most commonly in an auditory oddball paradigm in which a sequence of repetitive standard stimuli is interrupted infrequently and unexpectedly by a physically deviant 'oddball' stimulus. Electro- and magnetoencephalographic dipole mapping studies have localized the generators of MMN to supratemporal auditory cortex in the vicinity of Heschl's gyrus, but have not determined the degree to which MMN reflects activation within primary auditory cortex (AI) itself. The present study, using moveable multichannel electrodes inserted acutely into superior temporal plane, demonstrates a significant contribution of AI to scalp-recorded MMN in the monkey, as reflected by greater response of AI to loud or soft clicks presented as deviants than to the same stimuli presented as repetitive standards. The MMN-like activity was localized primarily to supragranular laminae within AI. Thus, standard and deviant stimuli elicited similar degrees of initial, thalamocortical excitation. In contrast, responses within supragranular cortex were significantly larger to deviant stimuli than to standards. No MMN-like activity was detected in a limited number to passes that penetrated anterior and medial to AI. AI plays a well established role in the decoding of the acoustic properties of individual stimuli. The present study demonstrates that primary auditory cortex also plays an important role in processing the relationships between stimuli, and thus participates in cognitive, as well as purely sensory, processing of auditory information.

Intracranial neurosurgery guided by functional imaging

Neurosurgery on eloquent cortex entails important risks of functional deficits complicating aggressive lesion resection. In this study, advanced biomagnetic functional imaging of somatosensory and motor cortex combined with surface rendered magnetic resonance imaging displays including vascular anatomy were used in conjunction with a new nonintrusive intraoperative guided instrumentation system to resect a tumor in eloquent cortex. Intraoperative verification of the accuracy of pre-operative motor localization demonstrated highly accurate results comparing direct stimulation and noninvasive presurgical mapping. The applicability of surface rendered combined functional and anatomic maps of cortex is directly evident on comparison of preoperative computer images and intraoperative pictures. This combination of new technologies has a significant potential for reduced risk and improved outcome in neurosurgery of eloquent cortex.

Intrasubject reliability and validity of somatosensory source localization using a large array biomagnetometer

Neuromagnetic fields were evoked by tactile stimuli and detected with a multi-channel biomagnetometer through 72 independent repetitive measurements on a single subject. Each measurement consisted of a somatosensory evoked response (N = 256 stimuli) using a single probe placement. These fields were then analyzed for source localization using an equivalent current dipole model and demonstrated highly reliable localizations. The 3 major neuromagnetic somatosensory response components peaking at 35, 65 and 110 msec all localized to the same area of cortex. The relative contributions of intrinsic brain activity, habituation, probe placement, and choice of fiduciary points for headframe determination were quantified. Intrinsic factors were found to constitute the major source of inter-measurement error. Sources localized by magnetic source imaging (MSI) appeared valid relative to neuroanatomical estimation of the central fissure on MRI. Non-invasive presurgical biomagnetic localization of somatosensory cortex produces reliable and valid functional localizations which can be of potential value in risk assessment and may provide a useful guide for invasive functional mapping.

The auditory evoked sustained field: origin and frequency dependence

A sound lasting for several seconds is known to elicit a baseline shift in electrical and magnetic records. We have studied the dependence of the magnetic field distribution of this "per-stimulatory" sustained field (SF) on tone frequency. Tone bursts of 2 sec duration and 60 dB nHL intensity were presented to 11 subjects at varying interstimulus intervals between 5 and 7 sec. The carrier frequencies of 250, 1000 and 4000 Hz varied randomly from trial to trial. The field distributions obtained are consistent with the view that the auditory evoked sustained field activity originates in the supratemporal cortex. Differences in the locations of equivalent current dipoles of the SF from those of the M100 wave of the slow auditory evoked field are consistent across subjects. The SF source locations corresponding to stimulus frequencies over an extended frequency range are arranged in a tonotopic manner and support the idea that the sources of the M100 and the SF are current dipole sheets located on the superior surface of the primary auditory cortex.

Relationship of transient and steady-state auditory evoked fields

Transient and steady-state auditory evoked fields (AEFs) to brief tone pips were recorded over the left hemisphere at 7 different stimulus rates (0.125-39 Hz) using a 37-channel biomagnetometer. Previous observations of transient auditory gamma band response (GBR) activity were replicated. Similar rate characteristics and equivalent dipole locations supported the suggestion that the steady-state response (SSR) at about 40 Hz represents the summation of successive overlapping (10 Hz) middle latency responses (MLRs). On the other hand, differences in equivalent dipole locations and habituation effects suggest that the magnetically recorded GBR is a separate phenomenon which occurs primarily at low stimulus rates and is unrelated to either the magnetically recorded MRL or SSR.

Evoked magnetic responses of the human auditory cortex to minor pitch changes: localization of the mismatch field

The neuromagnetic source localizations of the auditory M100 and the mismatch field (MMF) were studied using a large-array biomagnetometer. Standard tones of 1000 Hz and deviant tones of 1050 Hz were delivered with 90% and 10% probability, respectively. Wave forms of the derived MMF were computed by examining difference wave forms between the responses to the deviants and the responses to the standards preceding (D-P) and following (D-F) the deviants as well as to all remaining standards (D-A). The subset of standards preceding the deviants was used for a more realistic comparison with the set of deviants (having the same number of epochs and a similar signal-to-noise ratio), while the subset of standards following the deviants served to answer the question whether those standards also elicit an MMF. The MMF deflections were compared with each other, with the "native" MMF occurring in response to the deviants, and with wave M100. (The MMF as it appears in the unprocessed response to the deviants was termed "native" for an easy distinction from the "derived" MMF.) Our results demonstrate a distinct MMF deflection, corresponding in latency to the simultaneously recorded fronto-central electrical MMN. Source analysis, using a single moving dipole model, showed the same spatial localization for the native MMF and for the different derived MMFs. The MMF source location turned out to be significantly anterior, medial and inferior relative to the sources of the M100. The present data also demonstrate that a minor frequency deviation may not activate measurably different M100 generators, yet be sufficient to trigger the nearby but spatially distinct mismatch generator.