Dynamics of brain activation during picture naming

The cerebral representation of language, deduced from observing patients with brain lesions and from stimulations and recordings performed during brain surgery, has been further clarified by recent positron emission tomography and functional magnetic resonance imaging measurements. We now expand this static view into the dynamics of cortical activation using the accurate spatiotemporal resolution of whole-head magnetoencephalography. During picture naming, the conversion from visual to symbolic representation progressed bilaterally from the occipital visual cortex towards temporal and frontal lobes. Overt naming elicited the most widespread cortical activation. Some language-related sites also reacted, though more weakly or after a longer delay, during covert naming and even passive viewing.

Magnetoencephalographic evaluation of children and adolescents with intractable epilepsy

Magnetoencephalographic (MEG) discharges were recorded with multichannel superconducting quantum interference device (SQUID) gradiometers in 13 young candidates for epilepsy surgery. The sources of epileptic activity were related to generators of somatosensory and auditory evoked cortical responses and projected on magnetic resonance imaging (MRI) scans. Seven subjects had restricted or regional MEG foci, located in the frontoopercular (1), sensorimotor (3), perisylvian (1), mesiotemporal (1), or temporooccipital cortex (1). The MEG foci in the 3 patients who underwent operation agreed with the intracranial findings. Findings in the other patients emphasize the need to collect further data to define the ultimate role of MEG in preoperative evaluation of epilepsy.

Cortical reactivity in progressive myoclonus epilepsy

We studied 4 patients with progressive myoclonus epilepsy (Unverricht-Lundborg disease; ULD). Somatosensory evoked fields (SEFs), auditory evoked fields (AEFs), and spontaneous activity over the somatomotor cortex were recorded with a 24-channel SQUID gradiometer. All patients had "giant" 20-45 msec median nerve SEFs at the first somatomotor cortex, with 2-6 times larger amplitudes than the healthy control subjects. Later deflections were not similarly enhanced. The dependence of SEF amplitudes on interstimulus interval (0.2-4 sec) and on successive ulnar-median nerve stimulation (stimulus interval 40 msec) was comparable to that in controls. Cortical AEFs were attenuated and delayed. In 3 patients, the spontaneous activity consisted of 6-8 Hz mu rhythm, which originated within 2 cm from the sources of SEFs and was abolished by clenching of the contralateral fist. Control subjects had major spectral peaks around 10 and 20 Hz. The SEF amplitudes and the strength of the 6-10 Hz mu correlated strongly, suggesting that some components of evoked and spontaneous activity obtain contributions from overlapping neuronal populations. The results imply that ULD is associated with thalamo-cortical hyperreactivity in the sensorimotor but not in the auditory system.

Use of a computerized brain atlas in magnetoencephalographic activation studies

A pilot study was carried out to test the feasibility of an adjustable computerized brain atlas, adapted to the individual anatomy for localizing current dipoles by means of magnetoencephalography (MEG). The atlas can be adapted to individual computed tomography (CT) or magnetic resonance (MR) images. Position information is transferred between these imaging methods and MEG using a stereotactic technique. For this purpose, a special non-magnetic helmet was designed to be used together with the ordinary head fixation system. It seems likely that the proposed combination of the brain atlas with MEG, CT and MRI methods will become a powerful tool in exploring different brain functions.

Responses of the human auditory cortex to changes in one versus two stimulus features

Neuromagnetic responses were recorded with a 24-SQUID magnetometer in two "oddball" experiments to determine whether mismatch responses to changes in single stimulus features are additive. In experiment 1, the one-feature deviants differed from standards in interstimulus interval (ISI) or frequency, and the two-feature deviants in both ISI and frequency. In experiment 2, deviants differed in duration, frequency, or both. All deviants evoked a mismatch field (MMF) with sources close to each other in the supratemporal auditory cortex. Except for the ISI deviants, the MMF sources were about 1 cm anterior to the source of the 100-ms response, N100m, to the standards. In the two experiments, MMFs obtained in response to the two-feature deviants resembled closely the sum of MMFs in response to one-feature deviants. The results suggest that the standards leave a multiple neuronal representation in the human auditory cortex. The particular neuronal traces of the representation react independently to changes in different features of sound stimuli.

Independence of steady-state 40-Hz response and spontaneous 10-Hz activity in the human auditory cortex

We recorded neuromagnetic signals from 5 subjects with a 24-channel planar SQUID gradiometer. Our aim was to search for any relationship between the 40-Hz steady-state responses and the spontaneous 10-Hz activity in the auditory cortex. Source strengths and locations of spontaneous activity were determined from spectral components of the data, and sources within the auditory cortex were determined by restricting the source location to a neighborhood of the auditory evoked response to single clicks. The characteristics of the 40-Hz steady-state responses were not simply related to those of the spontaneous 10-Hz activity, even within subjects, suggesting that these two oscillatory phenomena have independent generators.

Parietal epileptic mirror focus detected with a whole-head neuromagnetometer

Whole-head magnetoencephalographic recordings revealed two parietal epileptic foci in homotopic areas of the hemispheres. The discharges occurred 17-20 ms later on the left than on the right hemisphere, implying the existence of a left-sided mirror focus. The foci were about 1 cm posterior to the hand primary somatosensory area, identified by evoked response measurements, and thus suggested epileptic activity at the parietal association cortex, in agreement with the observed callosal conduction time.

Magnetoencephalography Reveals Functions of the Human Brain

Magnetoencephalographic signals, detected noninvasively outside the human head, arise from intracellular currents in the fissural cortex. Localization of these currents gives spatially and temporally accurate information about functional organization of the healthy and diseased human cerebral cortex and about neural basis of cognitive functions.

Phonetic invariance in the human auditory cortex

Neuromagnetic signals evoked by synthesized syllables (/bae/ and /gae/) were recorded over the left auditory cortex of healthy humans. The fundamental frequencies of the syllables varied as if the same speaker had pronounced them at 16 different pitches. Specific mismatch responses to infrequent syllables among frequent syllables of the other type indicated that phonetically invariant information had been extracted at the level of the auditory cortex from the extensive irrelevant pitch variation. Such a detection mechanism is necessary for perceiving speech sounds in natural situations with a great deal of acoustic variation present.

Determinants of the auditory mismatch response

The auditory mismatch field (MMF) is supposed to reflect a comparison process between an infrequent deviant stimulus and the memory trace left by frequent standard stimuli. Therefore, the MMF amplitude has been thought to depend on the strength of such a trace. We examined this hypothesis in records with a 24-channel planar SQUID magnetometer by varying the number of stimuli preceding each deviant, the interdeviant interval (IDI) and the interstimulus interval (ISI) just preceding the deviant (pISI). When a constant IDI was employed and the number of standards between two deviants varied in different sessions, MMF amplitude increased as the number of standards increased. However, MMF did not depend on the number of standards between two deviants when the number varied within a single session and ISI varied as well. MMF decreased slightly when pISI increased from 0.6 to 3.4 sec. When IDI increased and the ISI remained constant, MMF amplitude increased. Most results can be explained within the framework of the memory-trace hypothesis of MMF generation. However, the strengthening of the trace seems to be a complex process which is also affected by the temporal features of the stimulus sequence.

Temporal integration and oscillatory responses of the human auditory cortex revealed by evoked magnetic fields to click trains

We recorded neuromagnetic evoked responses from the right auditory cortex of 7 healthy adults with a 24-channel planar SQUID gradiometer. The stimuli were 200-ms click trains presented at rates of 40, 80, 160 and 320 Hz, with interstimulus intervals (ISIs) of 1 and 4 s. The transient N100m response to the train onset depended on the click rate: the peak latency shortened to the same extent as the interval between successive clicks decreased in trains with rates from 40 Hz to 320 Hz. The N100m amplitude increased simultaneously, saturating at rates of 160-320 Hz. The mean N100m latency was slightly longer with the 1-s than with the 4-s ISI for all click rates. The systematic changes of the N100m amplitude and latency according to click rate demonstrate the importance of temporal integration for N100m generation, and imply an integration time of 20-25 ms. The 20- and 40-Hz click trains also elicited oscillatory 40-Hz responses 80-250 ms after the train onset. The 40-Hz responses were more resistant than N100m to changes of the ISI, and their sources slightly differed from those of N100m. These two responses evidently reflect different aspects of auditory processing.

Functional organization of the human first and second somatosensory cortices: a neuromagnetic study

Multichannel neuromagnetic recordings were used to differentiate signals from the human first (SI) and second (SII) somatosensory cortices and to define representations of body surface in them. The responses from contralateral SI, peaking at 20-40 ms, arose mainly from area 3b, where representations of the leg, hand, fingers, lips and tongue agreed with earlier animal studies and with neurosurgical stimulations and recordings on convexial cortex in man. Representations of the five fingers were limited to a cortical strip of approximately 2 cm in length. Responses from SII peaked 100-140 ms after contra- and ipsilateral stimuli and varied considerably from one subject to another. Signs of somatotopical organization were seen also in SII. Responses of SII were not fully recovered at interstimulus intervals of 8 s.

Human auditory cortical mechanisms of sound lateralization: II. Interaural time differences at sound onset

Neuromagnetic responses were recorded over the right temporal cortex using a 24-channel gradiometer. Stimuli were binaural click trains, presented with six separate interaural time differences (ITDs). N100m to sound onset was larger and earlier for stimuli presented with left- than with right-leading ITDs. With stimulus lateralization taken into account, monaural and binaural stimuli evoked responses of roughly equal amplitude. In selective adaptation and oddball experiments, stimuli presented with different ITDs excited overlapping neuronal populations, but the amount of overlap decreased as the ITD between the stimuli increased. There were no systematic differences in the cortical source locations of the N100m as a function of ITD, however. Thus it appears that ITD-sensitive neurons in the human auditory cortex are not organized into a large-scale, orderly representation, which could be resolved by MEG.

Human auditory cortical mechanisms of sound lateralization: I. Interaural time differences within sound

Neuromagnetic responses to 600-ms binaural click trains, presented once every 1.1 s, were recorded with a 24-channel gradiometer from 6 healthy humans. During the first 300 ms, the left-ear stimulus led the right by 0.7 ms and the sound was lateralized to the left ear. At 300 ms, the interaural time difference (ITD) changed and the lateralization moved to one of 5 different locations between the ears. An N100m response peaked about 110 ms after the sound onset and an N130mc response (c to stress a response to the change) about 135 ms after the ITD change. The source locations of N100m and N130mc agreed with activity in the supratemporal auditory cortex; this was confirmed in one subject by superimposing MEG results on MR images. The sources of N100m and N130mc did not differ statistically significantly from each other, nor were there differences in N130mc sources to various lateralization changes. N130mc grew larger when the ITD change increased, in parallel with the increase in the change of the perceived location. We suggest that N130mc is analogous to N100m, but is delayed due to postmasking induced by the early part of the sound.

Auditory evoked fields covary with perceptual grouping

Abrupt acoustic events evoke a transient magnetic response (N100m) at the supratemporal plane. Such responses decrease in amplitude as the interval between successive stimuli decreases to about 1 s. However, when pairs of stimuli are separated by still smaller intervals the second stimulus evokes a larger response than the first. This enhancement depends on the duration of the pause between the offset of the first stimulus and onset of the second. The range over which enhancement of N100m is observed agrees quite well with the range over which subjects experience perceptual grouping of the two stimuli with loudness enhancement of the second. Recordings from multi-channel SQUID gradiometers show that the effect involves not only a change in source strength but also a change in source location within supratemporal cortex. The results suggest that inhibition induced by onset of the first stimulus may be disinhibited by its offset. Physiological and psychological implications are discussed.

Cortical somatosensory magnetic responses in multiple sclerosis

Somatosensory evoked magnetic fields (SEFs) to contralateral median and ulnar nerve stimulation were analyzed in 10 patients with multiple sclerosis and in 8 healthy controls. SEFs were recorded with a 24-channel SQUID gradiometer over both hemispheres. Seven patients showed abnormally large-amplitude SEF deflections at 60-80 msec; 5 of them had multiple lesions around lateral ventricles in magnetic resonance imaging. In 2 patients with plaques at the level of 3rd and 4th ventricles and medulla, the 30 msec responses were enlarged. The equivalent sources of 20 msec and 30-80 msec responses were in the primary hand sensorimotor cortex both in patients and in control subjects. The results suggest that early and middle-latency SEFs reflect parallel processing of somatosensory input. Recording of middle-latency evoked responses, electric or magnetic, may give additional information about the somatosensory function in multiple sclerosis.

Magnetoencephalographic localization of epileptic cortex--impact on surgical treatment

A 24-channel, planar, superconducting quantum interference device gradiometer, sampling a fourth of the head surface over brain tissue, was used to determine the site of an epileptic focus in a 36-year-old woman with intractable complex partial epilepsy. The other presurgical findings appeared divergent: a large arachnoid cyst over the right parietal convexity, dissimilar interictal electroencephalographic patterns, and several neuropsychological dysfunctions. The equivalent current sources of magnetoencephalographic spikes were in the right posterior temporal region of the cortex, 4 cm apart from the cyst. Surgical exploration of the area pinpointed by magnetoencephalography revealed a pachygyric patch of cortex displaying focal discharges on the electrocorticogram. After resection, a dramatic reduction of seizures occurred. The good agreement between electrocorticography and magnetoencephalography warrants future investigation of multichannel magnetoencephalography as a potential alternative to invasive presurgical recordings.

Interaction of afferent impulses in the human primary sensorimotor cortex

We recorded somatosensory evoked magnetic fields (SEFs) over the hand area of the primary sensorimotor cortex (SMI) in 6 healthy adults in 2 sets of experiments to study interaction of afferent impulses. In experiment 1, SEFs were elicited by contralateral median nerve (MC) stimuli presented alone and 40 msec after a conditioning stimulus to the contralateral ulnar (UC), ipsilateral median (MI) or contralateral tibial (TC) nerve. N20m, P30m and P60m deflections to MC stimulation were markedly attenuated by preceding UC stimulation whereas N40m was enhanced, and a novel P80m emerged. In contrast, MI or TC stimulation did not affect the responses to MC. In experiment 2, the time course of recovery of N20m to median nerve stimuli was studied after stimulation of the adjacent ulnar and of the same median nerve. The recovery curves were similar for both conditioning stimuli with nearly full recovery of N20m at 120 msec. The results indicate marked interaction of impulses from ipsilateral median and ulnar nerves in human SMI, but no evidence was found of interaction from the two hands or from ipsilateral hand and foot.

Generator sites of spontaneous MEG activity during sleep

We have recorded spontaneous magnetoencephalographic (MEG) activity during overnight natural sleep in 4 healthy adults with a 24-channel SQUID gradiometer, mainly over the sides of the head. All sleep stages were obtained. The MEG wave forms resembled the EEG phenomena recorded simultaneously from the scalp midline, but the electric and magnetic signals did not always coincide. The source locations of different signals were studied by using a current dipole model. The equivalent sources of magnetic transients, resembling and often coinciding with the electric vertex waves and K-complexes, as well as the transients during REM sleep, were concentrated within a volume of 4 x 4 x 3 cm3 in the inferior parietal lobe. For spindles and slow waves, no such focal generators were found.

Modified activity of the human auditory cortex during auditory hallucinations

Previous reports have shown abnormalities in brain metabolism and evoked responses of schizophrenic patients with hallucinations. The authors recorded electric and magnetic auditory responses during transitory auditory hallucinations in two patients. Small but replicable response delays occurred during hallucinations. The results suggest that the effect of hallucinations on auditory cortex activity is similar to the effect of real sounds.

Neuromagnetic mismatch fields to single and paired tones

In 8 subjects we recorded multichannel magnetic responses to pitch changes in single 50 msec tones and in tone pairs (oddball paradigm; interstimulus interval 745 msec). Either "A" (1 kHz) was standard (90%) and "B" (1.2 kHz) deviant (10%), or "AA" was standard and "AB" deviant in pairs with onset asynchrony of 75 msec. Subject did not pay attention to the tones. A mismatch field (MMF) was evident in responses to deviants with an equivalent source in the supratemporal auditory cortex, about 1 cm anterior to that for N100m. The MMF source was 3-fold stronger for tone pairs than for single tones.

Neuromagnetic auditory evoked responses after a stroke in the right temporal lobe

Auditory evoked magnetic fields were recorded with a 24-SQUID magnetometer on a patient with a vascular lesion in the right auditory cortex and its vicinity. The contralateral transient 100 ms response (N100m) was absent over the damaged side, with no recovery within 16 months. Responses were normal over the healthy side. The responses to a binaural click train where interaural delays caused the perception of a shifting sound source; responses were again normal over the left side. Ipsilateral tones elicited N100m over the normal hemisphere but with smaller amplitude and longer latency than in healthy subjects. We suggest that ipsilateral responses are triggered by ipsilateral afferents but are significantly modified by excitatory callosal connections which did not work in our patient.