500-1000 Hz responses in the somatosensory system: approaching generators and function

Towards non-invasive multi-unit spike recordings: mapping 1kHz EEG signals over human somatosensory cortex

OBJECTIVE

Scalp-derived human somatosensory evoked potentials (SEPs) contain high-frequency oscillations (600 Hz; 'sigma-burst') reflecting concomitant bursts of spike responses in primary somatosensory cortex that repeat regularly at 600 Hz. Notably, recent human intracranial SEP have revealed also 1 kHz responses ('kappa-burst'), possibly reflecting non-rhythmic spiking summed over multiple cells (MUA: multi-unit activity). However, the non-invasive detection of EEG signals at 1 kHz typical for spikes has always been limited by noise contributions from both, amplifier and body/electrode interface. Accordingly, we developed a low-noise recording set-up optimised to map non-invasively 1 kHz SEP components.

METHODS

SEP were recorded upon 4 Hz left median nerve stimulation in 6 healthy human subjects. Scalp potentials were acquired inside an electrically and magnetically shielded room using low-noise custom-made amplifiers. Furthermore, in order to reduce thermal Johnson noise contributions from the sensor/skin interface, electrode impedances were adjusted to ≤ 1 kΩ. Responses averaged after repeated presentation of the stimulus (n=4000 trials) were evaluated by spatio-temporal pattern analyses in complementary spectral bands.

RESULTS

Three distinct spectral components were identified: N20 (<100 Hz), sigma-burst (450-750 Hz), and kappa-burst (850-1200 Hz). The two high-frequency bursts (sigma, kappa) exhibited distinct and partially independent spatiotemporal evolutions, indicating subcortical as well as several cortical generators.

CONCLUSIONS

Using a dedicated low-noise set-up, human SEP 'kappa-bursts' at 1 kHz can be non-invasively detected and their scalp distribution be mapped. Their topographies indicate a set of subcortical/cortical generators, at least partially distinct from the topography of the 600 Hz sigma-bursts described previously.

SIGNIFICANCE

The non-invasive detection and surface mapping of 1 kHz EEG signals presented here provides an essential step towards non-invasive monitoring of multi-unit spike activity.

Somatosensory evoked potentials recorded from the human pedunculopontine nucleus region

The pedunculopontine nucleus region (PPNR) is an integral component of the midbrain locomotor region and has widespread connections with the cortex, thalamus, brain stem, cerebellum, spinal cord, and especially, the basal ganglia. No previous study examined the somatosensory connection of the PPNR in human. We recorded somatosensory evoked potentials (SEP) from median nerve stimulation through deep brain stimulation (DBS) electrodes implanted in the PPNR in 8 patients (6 with Parkinson's disease, 2 with progressive supranuclear palsy). Monopolar recordings from the PPNR contacts showed triphasic or biphasic potentials. The latency of the largest negative peak was between 16.8 and 18.7 milliseconds. Bipolar derivation revealed phase reversal with median nerve stimulation contralateral to the DBS electrode in 6 patients. There was no difference in SEP amplitude and latency between on and off medication states. We also studied the high frequency oscillations (HFOs) by filtering the signal between 500 and 2,500 Hz. The HFOs could be identified only from contralateral stimulation and had intraburst frequencies of 1061 ± 121 Hz, onset latencies of 13.8 ± 1.2 milliseconds, and burst durations of 7.3 ± 1.1 milliseconds. Among the 10 recordings with HFOs, only 1 had possible phase reversal in the bipolar derivation. Our results suggest that there are direct somatosensory inputs to the PPNR. The slow components and HFOs of the SEP have different origins.

Detection of seizure-associated high-frequency oscillations above 500Hz

High-frequency oscillations (HFOs) of up to 500Hz in EEG are considered to have close relation with ictogenesis. We had the unique opportunity to record a seizure in EEG with intracerebral macroelectrodes and a sampling frequency of 10kHz. Considering the notion that faster HFOs are likely more ictogenic, we investigated this ictal EEG data to find if even faster HFOs were present. HFOs were investigated in interictal spikes and seizure activity using time-frequency spectra: t values corresponding to frequencies from 100 to 1000Hz were obtained by comparison to the background and controlled by the false discovery rate (FDR). The seizure had a right hippocampal onset. HFOs up to 800Hz as well as HFOs below 500Hz built up in the hippocampal discharges more at the beginning of the seizure and during the preictal period than in the interictal period. These HFOs were visually confirmed in temporally expanded EEG traces. We demonstrated for the first time the existence of HFOs above 500Hz and up to 800Hz with intracerebral macroelectrodes in an epileptic patient; they occurred primarily in association with the seizure discharge. HFOs above 500Hz possibly reflect facilitation of ictogenic neuronal hypersynchronization.