Nonlinear interactions of high-frequency oscillations in the human somatosensory system

OBJECTIVE: The source of somatosensory evoked high-frequency activity at about 600 Hz is still not completely clear. Hence, we aimed to study the influence of double stimulation on the human somatosensory system by analyzing both the low-frequency activity and the high-frequency oscillations (HFOs) at about 600 Hz. METHODS: We used median nerve stimulation at seven interstimuli intervals (ISIs) with a high time resolution between 2.4 and 4.8 ms to investigate the N15, N20 and superimposed HFOs. Simultaneously, the electroencephalogram and the magnetoencephalogram of 12 healthy participants were recorded. Subsequently, the source analysis of precortical and cortical dipoles was performed. RESULTS: The difference computations of precortical dipole activation curves showed in both the low- and high-frequency range a correlation between the ISI and the latency of the second stimulus response. The cortical low-frequency response showed a similar behavior. Contrarily, in the second response of cortical HFOs this latency shift could not be confirmed. We found amplitude fluctuations that were dependent on the ISI in the low-frequency activity and the HFOs. These nonlinear interactions occurred at ISIs, which differ by one full HFO period (1.6 ms). CONCLUSIONS: Low-frequency activity and HFOs originate from different generators. Precortical and cortical HFOs are independently generated. The amplitude fluctuations dependent on ISI indicate nonlinear interference between successive stimuli. SIGNIFICANCE: Information processing in human somatosensory system includes nonlinearity.     

Decrement of N20 amplitude of the median nerve somatosensory evoked potential in Creutzfeldt-Jakob disease patients.

We studied somatosensory evoked potentials (SSEPs) in eight Creutzfeldt-Jakob disease (CJD) patients presenting with subacute progressive dementia, generalized myoclonus, and characteristic periodic sharp wave complexes in EEG. Somatosensory evoked potentials were elicited by median nerve stimulation at the wrist. We compared SSEP findings with EEG and the clinical stage proposed by the Japanese Slow Virus Infection Research Committee (stage 1: early stage to stage 5: terminal stage). Until clinical stage 3, short-latency SSEPs showed normal findings despite the severely abnormal EEG. With the progression to clinical stages 4 and 5, however, the amplitude of N20 began to decrease and finally disappeared without prolongation of the latency of N20, whereas other short-latency components were preserved. We recorded giant SSEPs in two of three patients in stage 4, when the periodic sharp wave complex in EEG began to decrease in amplitude. The giant SSEPs decreased in amplitude with the progression of the illness. These findings suggest that the short-latency SSEP is relatively preserved until the middle phase of the disease but that it is eventually affected in the terminal phase. We conclude that our results are compatible with the CJD pathologic findings and that the amplitude of N20 reflects the extent of cortical damage in CJD patients.     

High-frequency somatosensory thalamocortical oscillations and psychopathology in schizophrenia

Human cortical somatosensory evoked potentials (SEPs), which are presumably generated in afferent thalamocortical and early cortical fibers, reveal a burst of superimposed early (N20) high-frequency oscillations (HFOs), around 600 Hz. There is increasing evidence of an imbalance of thalamocortical systems in schizophrenic patients. In order to assess correlations between somatosensory evoked oscillations and symptoms of schizophrenia, we investigated median nerve SEPs in 20 inpatients and their age-matched and gender-matched healthy controls using a multichannel EEG. Dipole source analysis and wavelet transformation were performed before and after application of a 450-Hz high-pass filter. In schizophrenics, the maximum HFOs occurred with a significantly prolonged latency. There was also a higher amplitude (energy) in the low-frequency range of the N20 component compared with the controls. Importantly, amplitudes (energy) of HFOs were inversely correlated with symptoms of formal thought disorder and delusions. Alterations of the thalamocortical somatosensory signal processing in schizophrenia with absence of an early HFO - assumed to be of inhibitory nature - could indicate a dysfunctional thalamic inhibition with increased amplitudes of N20, paralleled by enhanced positive schizophrenic symptoms.     

Somatosensory evoked high-frequency oscillations in migraine patients.

OBJECTIVE: To explore additional evidence concerning generators of somatosensory evoked high-frequency oscillations (HFOs). METHODS: We recorded HFOs in migraine patients. Subjects were 19 healthy normal subjects and 19 migraineurs. Electrical stimuli were delivered alternately to the right and left median nerves at their wrists. EEGs were recorded from C3'-Fz, C4'-Fz, Erb1-Erb2, Erb2-Erb1 and Cv6-Fz using a 0.3 Hz low-frequency filter and a 3000 Hz high-frequency filter. Responses to 5000 stimuli were averaged. For separation of HFOs from underlying N20, the digitized wide-band signals were digitally bandpass filtered (400-800 Hz) and averaged. RESULTS: There were no significant differences in peak latencies and amplitudes for N9, N13, N20 and P25 components between normal controls and migraineurs. Root-mean-square amplitudes for HFOs in migraineurs were significantly diminished compared with normal controls. CONCLUSIONS: A diminished inhibitory mechanism in the somatosensory system may exist in migraineurs. It remains to determine what cell populations contribute to generating HFOs. SIGNIFICANCE: This indicates that there is a dysfunction in cortical information processing in the somatosensory cortex of migraineurs.     

Patterns of disturbed impulse propagation in multiple sclerosis identified by low and high frequency somatosensory evoked potential components.

In human median nerve somatosensory evoked potentials (SSEPs), high frequency (600 Hz) oscillations (HFOs) are superimposed onto the low frequency SSEP component N20. High frequency oscillations are generated both in deep axon segments of thalamo-cortical projection neurons and at the primary somatosensory cortex. The present study aimed to test the hypothesis that HFOs might be more sensitive to temporal dispersion caused by demyelinating lesions in multiple sclerosis (MS) than the N20. The authors recorded HFOs in median nerve SSEPs in 50 patients with definite MS and in 30 healthy controls. Three patterns of SSEP alterations were found: (1) abolished HFOs with either normal (11% of stimulated limbs), or delayed N20 (16% of stimulated limbs); (2) an attenuation of N20 amplitude with preserved HFOs (13%); and (3) a mixture of both patterns (21%). The first pattern--normal N20 with abolished HFOs--indicates that the HFOs are a sensitive marker of slight demyelination. The second pattern is suggestive of a mainly axonal lesion type, while the third pattern points to a combined axonal/demyelinating process or a conduction block. Analysis of HFOs allows identification of slight demyelinating processes in MS patients in whom the N20 SSEP component remains unaffected. The HFOs provide a tool to distinguish different patterns of disturbed impulse propagation.     

Epileptic spasms in older pediatric patients: MEG and ictal high-frequency oscillations suggest focal-onset seizures in a subset of epileptic spasms

OBJECTIVES: To elucidate the pathophysiology of intractable epileptic spasms in older children by describing the interictal magnetoencephalography spike sources (MEGSSs), intracranial EEG ictal-onset zones (IOZs) and their ictal high-frequency oscillations (HFOs) and surgical outcomes. METHODS: We studied five patients (4.5-14 years) who underwent surgery following intracranial video-EEG (VEEG) monitoring. We analyzed clinical profiles, MRIs, scalp and intracranial VEEGs, and MEGSSs. We localized ictal HFOs using a sampling rate of 1000 Hz and multiple band frequency analysis (MBFA). RESULTS: Seizure onset ranged from 0.4 to 8 years. Three patients presented with asymmetrical spasms. Interictal scalp VEEG recorded predominantly unilateral epileptiform discharges in four; generalized and multifocal in one. Ictal scalp VEEG showed generalized high-amplitude slow waves with superimposed fast waves in four patients; hemispheric electrodecremental episodes in one. MRI findings were normal in three, hemispheric polymicrogyria and periventricular heterotopia in one each. All patients had unilateral MEGSS clusters. Ictal HFOs, ranging from 150 to 250 Hz, localized over Rolandic and frontal regions in four, with one also having extensive temporo-occipital HFOs. After cortical resection, three patients were seizure free; one had >90% reduction in seizure frequency. One patient experienced residual seizures after resection of the hemispheric ictal HFO region. CONCLUSION: Unilateral clusters of MEGSSs overlapped regional IOZs in older patients with epileptic spasms. High spatio-temporal MBFA before and during spasms revealed the regional ictal HFOs. Seizure-free outcomes following resection of zone with MEGSS clusters and ictal HFOs suggested that a subset of epileptic spasms was focal-onset seizures.     

Auditory stimulation enhances thalamic somatosensory high‐frequency oscillations in healthy humans: a neurophysiological marker of cross‐sensory sensitization?

Electrical stimulation of upper limb nerves evokes a train of high‐frequency wavelets (high‐frequency oscillations, HFOs) on the human scalp. These HFOs are related to the influence of arousal‐promoting structures on somatosensory input processing, and are generated in the primary somatosensory cortex (post‐synaptic HFOs) and the terminal tracts of thalamocortical radiations (pre‐synaptic HFOs). We previously reported that HFOs do not undergo habituation to repeated stimulations; here, we verified whether HFOs could be modulated by external sensitizing stimuli. We recorded somatosensory evoked potentials (SSEPs) in 15 healthy volunteers before and after sensitization training with an auditory stimulus. Pre‐synaptic HFO amplitudes, reflecting somatosensory thalamic/thalamocortical activity, significantly increased after the sensitizing acoustic stimulation, whereas both the low‐frequency N20 SSEP component and post‐synaptic HFOs were unaffected. Cross‐talk between subcortical arousal‐related structures is a probable mechanism for the pre‐synaptic HFO effect observed in this study. We propose that part of the ascending somatosensory input encoded in HFOs is specifically able to convey sensitized inputs. This preferential involvement in sensitization mechanisms suggests that HFOs play a critical role in the detection of potentially relevant stimuli, and act at very early stages of somatosensory input processing.     

Short-latency somatosensory evoked potential and event-related potential in patients with multiple cerebral infarcts.

Although the short-latency somatosensory evoked (SSEPs) and the cognitive event-related potentials (ERPs) have been found to change in patients with cerebral vascular disease, the relationship between these parameters has yet to be determined. For clarification of this relationship, SSEPs and ERPs were measured in 33 patients with multiple cerebral infarcts (MCI) and 25 age-matched normal subjects. ERPs were recorded during auditory discrimination tasks. The latency of P300 from the Pz region was measured. SSEPs evoked by median nerve stimulation were recorded from the second cervical vertebra and contralateral primary somatosensory cortex with a midfrontal reference. The central conduction time (CCT), the interpeak latency between N13 and N20, was measured. P300 latency in patients with MCI was significantly longer than that of the normal subjects. Patients with MCI also showed longer CCT than the normal subjects. CCT and P300 latency were significantly correlated in patients with MCI. This correlation was not found in normal subjects. From these results, the severity of the lesion with respect to SSEPs appears related to the prolongation of P300 latency in patients with MCI.     

Somatosensory evoked potentials as indicators of altered cerebral excitability during psychotropic drug treatment

We observed an increase in the amplitude of the early cortical somatosensory evoked potentials (SSEPs) in five patients who developed myoclonus and/or generalized seizures during treatment with antidepressants. The increases correlated closely with the course of the clinical disturbances. In every case the SSEPs returned to normal values after the discontinuation of the psychotropic drugs. We suggest that SSEPs might help to identify and monitor patients who are at an increased risk of potentially hazardous side effects during psychopharmacological treatment.     

Influence of cholinergic circuitries in generation of high-frequency somatosensory evoked potentials.

OBJECTIVE: High-frequency oscillations (HFOs) evoked by upper limb stimulation reflect highly synchronised spikes generated in the somatosensory human system. Since acetylcholine produces differential modulation in subgroups of neurons, we would determine whether cholinergic drive influences HFOs. METHODS: We recorded somatosensory evoked potentials (SEPs) from 31 scalp electrodes in 7 healthy volunteers, before and after single administration of rivastigmine, an inhibitor of central acetylcholinesterase. Right median nerve SEPs have been analysed after digital narrow bandpass filtering (500-700 Hz). Raw data were further submitted to Brain Electrical Source analysis (BESA) to evaluate the respective contribution of lemniscal, thalamic and cortical sources. Lastly, we analysed by Fast Fourier transform spectral changes after drug administration in the 10-30 ms latency range. RESULTS: Rivastigmine administration caused a significant increase of HFOs in the 18-28 ms latency range. Wavelets occurring before the onset latency of the conventional N20 SEP did not show any significant change. A similar increase concerned the strength of cortical dipolar sources in our BESA model. Lastly, we found a significant power increase of the frequency peak at about 600 Hz in P3-F3 traces after drug intake. CONCLUSIONS: Our findings demonstrate that the cortical component of HFOs is significantly enhanced by cholinergic activation. Pyramidal chattering cells, which are capable to discharge high-frequency bursts, are mainly modulated by cholinergic inputs; by contrast, acetylcholine does not modify the firing rate of fast-spiking GABAergic interneurons. We thus discuss the hypothesis that cortical HFOs are mainly generated by specialised pyramidal cells.     

High frequency oscillations in the somatosensory evoked potentials (SSEP's) are mainly due to phase-resetting phenomena

A small series of high frequency oscillations (HFOs) overlapping the earliest part of the N20 wave can be observed in the somatosensory evoked potentials (SSEPs) of normal subjects. We tried to elucidate whether these high frequency components are mainly due to phase-resetting phenomena, to the emergence of new oscillations related to the stimuli, or to a combination of both. Averaged median-nerve SSEPs from seven healthy subjects were studied by means of time-frequency analysis. The presence of new oscillatory activities was evaluated by averaging the energy of the single-trial time-frequency transforms in the HFOs range (400-1000 Hz). To study phase-resetting phenomena, we measured inter-trial coherence (ITC) in the same frequency range. A marked inter-trial coherence related to the HFOs was found, whereas energy changes (related to the emergence of new oscillations) were minimal. The combination of these three different approaches suggests that the HFOs are mainly due to resettings of the ongoing EEG activity originated in response to the stimuli. The emergence of new activities does not seem to be a relevant mechanism in the formation of these components.     

Cortical myoclonus in childhood and juvenile onset Huntington's disease

ObjectiveHuntington's disease (HD) appearing before the age of 20 years gives rise to a distinct phenotype with respect to the classical adult-onset disease. Here we describe three patients with childhood or juvenile HD onset presenting with action myoclonus.MethodsWe performed jerk-locked back-averaging (JLBA), EEG-EMG coherence and phase analysis, long-loop reflexes (LLRs) and somatosensory evoked potentials (SSEPs). In one patient, we also performed transcranial magnetic stimulation (TMS) using single and paired pulses.ResultsIn all patients, the EMG features revealed movement activated quasi-rhythmic repetitive jerks; the JLBA and EEG-EMG spectral and coherence profiles indicated a cortical generator of the myoclonus. All patients had enhanced LLRs during muscle contraction, while none showed giant SSEPs. The evaluation of intracortical inhibition by means of TMS revealed reduced inhibition at short and long interstimulus intervals.ConclusionsThe rhythmic course of the action myoclonus and the characteristics of the LLRs suggest that myoclonus is due to a reverberant circuit involving the motor cortex, possibly because of an imbalance between excitatory and inhibitory cortical neuronal systems.SignificanceOur findings suggest a similar cortical dysfunction in childhood and juvenile onset HD, which probably results from a specific circuitry impairment.     

Methyl bromide myoclonus: an electrophysiological study

We report a case of myoclonus from overnight exposure to methyl bromide. Myoclonus was either spontaneous or induced by somatosensory stimulation or voluntary movements, multifocal and sometimes generalized. Median SEP showed normal size P14-N20, but giant parietal P25, N33 and frontal P22-N30 waves. Back-averaging showed a biphasic EEG spike of maximal amplitude at the central region contralateral to the corresponding myoclonic jerk recorded from abductor pollicis brevis and preceding it by a short interval consistent with conduction in corticospinal pathways. Long latency reflexes from cutaneous and mixed nerve stimulation were enhanced. The above electrophysiological findings suggest that myoclonus following methyl bromide poisoning belongs to the cortical reflex myoclonus category.     

Rare cause of myoclonus with giant SEP's: methyl bromide poisoning. Apropos of a case with unilateral predominance

Giant and asymmetric SEPs were recorded in a patient with predominantly unilateral, spontaneous and intention myoclonus due to voluntary intoxication with methyl bromide as soon as day 3 after intoxication. The N10 Erb's point potential, cervical N13 and scalp recorded P15 potentials were found to be normal in latency, morphology and amplitude. The somesthetic informations could be considered as normally processed up to the subcortical levels of the somatosensory pathways. The parietal cortical potentials N20 and P25 and the frontal cortical potentials P22 and N30, contralateral to myoclonus, were abnormally large. This suggests that myoclonus could be related with an abnormal reactivity of somatomotor and somatosensory cortices to the afferent volleys triggered by voluntary movements. The prerolandic components P22 and N30 were found to be relatively more enhanced than the parietal N20 and P15.     

Lack of activation of human secondary somatosensory cortex in Unverricht-Lundborg type of progressive myoclonus epilepsy

Previous electroencephalographic and magnetoencephalographic studies have demonstrated giant early somatosensory cortical responses in patients with cortical myoclonus. We applied whole-scalp magnetoencephalography to study activation sequences of the somatosensory cortical network in 7 patients with Unverricht-Lundborg-type progressive myoclonus epilepsy diagnostically verified by DNA analysis. Responses to electric median nerve stimuli displayed 30-msec peaks at the contralateral primary somatosensory cortex that were four times stronger in patients than in control subjects. The amplitudes of 20-msec responses did not significantly differ between the groups. In contrast to control subjects, 5 patients displayed ipsilateral primary somatosensory cortex activity at 48 to 61 msec in response to both left- and right-sided median nerve stimuli. Furthermore, their secondary somatosensory cortex was not significantly activated. These abnormalities indicate altered responsiveness of the entire somatosensory cortical network outside the contralateral primary somatosensory cortex in patients with Unverricht-Lundborg-type progressive myoclonus epilepsy. The deficient activation of the secondary somatosensory cortex in Unverricht-Lundborg patients may reflect disturbed sensorimotor integration, probably related to impaired movement coordination.     

Effects of general anesthesia on high-frequency oscillations in somatosensory evoked potentials.

To determine the characteristics of high-frequency oscillations (HFOs) of cortical somatosensory evoked potentials (SEPs), the effect of general anesthesia on HFOs and low-frequency primary cortical responses was studied. The authors recorded SEPs elicited by median nerve stimulation directly from human brains of seven patients who underwent implantation of subdural electrodes before surgical treatment of intractable epilepsy. Recordings were made before and during general anesthesia. Changes in the number of HFOs and amplitude ratios of HFOs/primary cortical responses were analyzed. Under general anesthesia, the number of HFO peaks and the amplitude ratios were significantly decreased. General anesthesia induced remarkably decreased HFO activities when compared to low-frequency SEPs, suggesting that each of those originated from different generators. Possible relations between gamma-amino-butyric acid (GABA)ergic inhibitory interneurons and HFOs are discussed.     

Motor responses to afferent stimulation in juvenile myoclonic epilepsy

PURPOSE: To document whether the mechanisms responsible for myoclonic jerks in juvenile myoclonic epilepsy (JME) are similar to those causing other forms of myoclonus. METHODS: We studied somatosensory evoked potentials, the conditioning effect of cutaneous afferents on motor potentials evoked by transcranial magnetic stimulation (TMS), and intracortical inhibition and facilitation in response to paired TMS in a group of nine patients with JME and 20 normal controls. RESULTS: Intracortical inhibition was abnormal, whereas cortical somatosensory evoked potentials and TMS conditioned by cutaneous afferents were unaltered in JME patients. CONCLUSIONS: Abnormal processing of cutaneous afferents would not appear to contribute to myoclonus in JME.     

Small-amplitude cortical myoclonus in Parkinson's disease: physiology and clinical observations.

We studied the occurrence of small-amplitude myo- clonus in 20 idiopathic Parkinson's disease patients who had no evidence of dementia as defined by criteria in the Diagnostic and Statistical Manual of Mental Disorders, fourth edition. Parkinson's disease was diagnosed by United Kingdom Brain Bank criteria, and clinical assessment was performed with the Unified Parkinson's Disease Rating Scale motor score, Hoehn and Yahr staging, and the Mini-Mental State Examination. Clinical assessment showed a range of mild-to-moderate disease severity. All patients underwent polygraphic electro-encephalographic-electromyographic (EMG) recording with back-averaging, somatosensory evoked potential testing, and attempted elicitation of long-latency EMG responses. Multichannel surface EMG recording during muscle activation showed irregular, multifocal, brief (<50 msec) myoclonus EMG discharges. Back-averaging consistently showed a focal, short-latency, electroencephalographic transient prior to the myoclonus EMG discharge. Cortical somatosensory evoked potential waves were not enlarged, and long-latency EMG responses at rest were not present. The small-amplitude myoclonus in such cases arises from an abnormal discharge from the sensorimotor cortex. The mechanism of this cortical myoclonus in Parkinson's disease has differences from the more common "cortical reflex myoclonus" physiology. Advanced parkinsonism is not a requirement for manifestation of this myoclonus type. Although the myoclonus occurred without dementia in these cases, its relationship to the subsequent development of cognitive impairment remains to be defined.     

High-frequency magnetic oscillations evoked by posterior tibial nerve stimulation

Magnetocephalographic recordings of the primary somatosensory response (P37m) and high-frequency oscillations (HFOs) evoked by posterior tibial nerve stimulation were obtained in normal subjects. Electrical stimuli were delivered to the posterior tibial nerve and magnetic recordings were taken over the superior aspect of the left hemisphere with a 37-channel biomagnetometer. In order to separate the high-frequency oscillations from the underlying P37m, the wide-band (0.1-1200 Hz) recorded responses were digitally filtered with a 500-800 Hz band-pass filter. The localization of the HFOs were estimated to be in somatosensory area 3b, very close to the P37m source. Our data suggest that the HFOs are somatotopically arranged in the primary somatosensory cortex, and are a ubiquitous phenomenon of the primary somatosensory cortex.     

Electrophysiological studies in cerebrotendinous xanthomatosis.

Seven patients with cerebrotendinous xanthomatosis (CTX) were studied by electrophysiological techniques. The percentages of abnormalities detected in nerve conduction studies and electroencephalograms were 28.6% (two patients) and 100%, respectively. All patients showed prolonged central conduction times in short latency somatosensory evoked potentials (SSEPs) by tibial nerve stimulation but normal SSEPs by median nerve stimulation. Brain stem auditory evoked potentials and visual evoked potentials were abnormal in three (42.9%) and four patients (57.1%), respectively. These electrophysiological parameters were correlated with the ratio of serum cholestanol to cholesterol concentration. The results of SSEPs suggest that the polyneuropathy in CTX is caused by distal axonopathy affecting longer axons before shorter axons (central-peripheral distal axonopathy).