Multichannel detection of magnetic compound action fields with stimulation of the index and little fingers

Can magnetoencephalography track the afferent information flow along white matter thalamo-cortical fibers?

White matter thalamo-cortical fibers allow the communication of distant brain regions by carrying neuronal signals. Mapping non-invasively the information flow within white matter fibers is regarded so far as impossible. We investigated here whether information flow propagating along thalamo-cortical fibers can be detected using magnetoencephalography (MEG). Somatosensory evoked fields (SEFs) were recorded from healthy subjects and a patient with a unilateral, prenatally acquired, white matter lesion, which had induced the development of an abnormal trajectory of thalamo-cortical fibers. Equivalent current dipole (ECD) was used to model sources of SEFs. ECD at ~15 ms after stimulus onset was located within or close to the contralateral thalamus at the proximity of a hemodynamic response detected during a similar fMRI experiment. At the M20 peak latency, ECD was localized within the hand area of the contralateral primary somatosensory cortex (Brodmann area 3b (BA3b)). In healthy subjects, ECD changed dynamically position from thalamus to BA3b following a curved path, which was partially overlapping the thalamo-cortical fibers reconstructed by tractography. In the patient, ECD followed a similar path only in the intact hemisphere. In the affected hemisphere, the dipole trajectory circumnavigated the extended lesion on its way to the preserved primary somatosensory cortex--similar to the trajectory findings. Evidence from different methodological approaches converges on the conclusion that MEG can track the afferent information flow along thalamo-cortical fibers and in contrast to the traditional view can localize under presuppositions deep thalamic sources.

Magnetoneurography: theory and application to peripheral nerve disorders

Magnetoneurography (MNG) is a non-invasive method to trace and visualize three-dimensionally the propagation path of compound action currents (CAC) along peripheral nerves. The basic physical and physiological principle is the mapping of extremely weak magnetic fields generated by the intraaxonal longitudinal ion flows of evoked nerval CAC using SQUID sensors (Superconducting Quantum Interference Devices). During recent years, MNG protocols have been established which allow for a non-invasive spatiotemporal tracing of impulse propagation along peripheral nerves in humans and in particular along proximal nerve segments in a clinical setting. Thereby, the three-dimensional path, the local nerve conduction velocity, the length and strength of the CAC de- and repolarization phase have been reconstructed. First recordings in patients demonstrated that the method is sensitive enough to detect and to localize nerve conduction anomalities along nerve roots, as, e.g. caused by lumbosacral disc herniation. This review on MNG will focus on those studies which provide data from humans and thereby reveal perspectives for its future clinical applications.

[Magnetoneurographic registration of evoked summation action fields over lumbar vertebrae following transcutaneous tibial nerve stimulation]

Goal of this study was the development of a protocol for the registration of evoked magnetic fields over the lumbar spine using off-the-shelf equipment. Three subjects in a sitting position with their torso bent slightly forward were stimulated at the tibial nerve with a commercially available stimulator. Neuromagnetic fields were registered over a circular, 800 cm2 area of the lumbosacral spine using a 61-channel 4D-Neuroimaging biomagnetometer. After appropriate signal processing, dipolar magnetic fields with a field strength 5-17 fT peak-to-peak amplitude were detected in three out of four registrations. Location and orientation of these fields concurred with the expected evoked compound action currents along the course of the nerve fibers.

High-frequency oscillations of somatosensory evoked potentials and fields

A short review of previous studies is presented on somatic, evoked high-frequency oscillations. Also described briefly is recent data on the detection of high-frequency oscillations to posterior tibial nerve stimulation, and also on both tangential (area 3b) and radial (area 1) dipoles to median nerve stimulation. The findings show that high-frequency oscillations are not specific to median nerve stimulation but represent ubiquitous activity in the primary somatosensory cortex. Modulation of high-frequency oscillations versus electric and magnetic N20, N20 (m), primary response by a wake-sleep cycle, by attention or interference, by aging, and in central nervous system diseases such as Parkinson's disease and myoclonus epilepsy are also presented. Finally, a gamma-aminobutyric acid inhibitory interneuron hypothesis is presented for high-frequency oscillations based primarily on the findings regarding reciprocal modulation of the high-frequency oscillations and the underlying magnetic N20 (N20m) by a wake-sleep cycle and by attention or interference.

Reciprocal modulation of somatosensory evoked N20m primary response and high-frequency oscillations by interference stimulation

OBJECTIVES

We examined whether the inverse relation between somatic evoked N20m primary response and high-frequency oscillations during a wake-sleep cycle (Hashimoto, I., Mashiko, T., Imada, T., Somatic evoked high-frequency magnetic oscillations reflect activity of inhibitory interneurons in the human somatosensory cortex, Electroenceph clin Neurophysiol 1996;100:189-203) holds for interference stimulation.

METHODS

Somatosensory evoked fields (SEFs) from 14 subjects were measured following electric median nerve stimulation at the wrist with, and without, concurrent brushing of the palm and fingers. SEFs were recorded with a wide bandpass (0.1-1200 Hz) and then N20m and high-frequency oscillations were separated by subsequent low-pass (< 300 Hz) and high-pass (> 300 Hz) filtering.

RESULTS

The N20m decreased dramatically in amplitude during interference stimulation. In contrast, the high-frequency oscillations moderately increased in number of peaks.

CONCLUSIONS

These results demonstrate the presence of an inverse relation between N20m and high-frequency oscillations for interference stimulation. We speculate that the high-frequency oscillations represent a localized activity of GABAergic inhibitory interneurons of layer 4, characterized by a high-frequency spike burst (200-1000 Hz) without adaptation, and that the continuous interference stimulation induces tonic excitation of the interneurons, leading to a facilitation of responses to the coherent afferent volley elicited by the median nerve stimulation (bottom-up mechanism). On the other hand, refractoriness of the pyramidal neurons caused directly by interference stimulation along with an enhanced feed-forward inhibition from the interneurons will lead to a decrease of N20m amplitude.

Magnetoneurographic 3D localization of conduction blocks in patients with unilateral S1 root compression

OBJECTIVES

Tibial nerve somatosensory evoked magnetic fields (tSEFs) over the lower back reflect the propagation of compound action currents along fibers of plexus, nerve roots and cauda equina. One clinical perspective for this 'magnetoneurography' is the non-invasive 3D localization of focal slowing or blocks of conduction. Here, first tSEF mappings in 3 consecutive patients with acute unilateral S1 nerve root compression are reported.

METHODS

Right and left tibial nerves were electrostimulated in alternation; tSEF responses were recorded using a multichannel SQUID-detector; additionally, spinal and cortical SEP, F-wave and H-reflex studies were performed.

RESULTS

In all patients an intraindividual side-to-side comparison of spinal tSEF mappings was obtained: using a dipolar source model compound action currents could be visualized propagating along plexus, nerve roots and cauda equina on the non-affected side whereas on the affected side normally-propagating dipolar field patterns could be recorded only distal to the spinal transforaminal root entrance; this reflects focal slowing or block of conduction in nerve root fibers as indicated by the SEP, F-wave and H-reflex study results.

CONCLUSIONS

With a registration time of 15 min a 3D localization of proximal slowing or block of conduction was successfully performed in patients suffering from acute nerve root lesions.

Mapping of tibial nerve evoked magnetic fields over the lower spine

Using a low-noise 49-channel dc-SQUID system spinal somatosensory evoked fields (SEF) were recorded which were generated by compound action currents evoked upon posterior tibial nerve stimulation. The SEF mapping showed the action current propagation along the sciatic nerve, lumbosacral plexus and cauda equina in parallel to simultaneously recorded electrical potentials (SEP). For a reliable intraindividual side-to-side comparison of spinal SEFs the right and left tibial nerves were stimulated in alternating order; this procedure minimizes artifactual inter-nerve SEF map differences due to eventual patient-to-sensor displacements which might occur in serial measurements. These large-area lumbar SEF mappings open up several clinical perspectives for magnetoneurography, in particular with respect to the 3D-localization of proximal conduction blocks.

Magnetometry of evoked fields from human peripheral nerve, brachial plexus and primary somatosensory cortex using a liquid nitrogen cooled superconducting quantum interference device

Superconducting Quantum Interference Devices (SQUIDs) can be used to detect neuromagnetic fields evoked in the peripheral and central nervous system. Up to now, such measurements had to be based on SQUIDs with a low critical temperature (Tc) requiring liquid helium cooling. Recent improvements in high-Tc SQUID technology relying on liquid nitrogen cooling led to a significant reduction in the system's noise level. Hare, first high-Tc recordings of weak neuromagnetic fields are demonstrated. In particular, along the entire somatosensory afferent pathway including peripheral nerves, brachial plexus and primary somatosensory neocortex evoked neuromagnetic activities were detected using conventional recording parameters for bandwidth and number of averages. This opens up a wide perspective for cost-effective high-Tc magnetometry in clinical neuroscience.