Cortical potentials evoked by epidural stimulation of the cervical and thoracic spinal cord in man

Scalp somatosensory evoked potentials (SEPs) were recorded after electrical stimulation of the spinal cord in humans. Stimulating electrodes were placed at different vertebral levels of the epidural space over the midline of the posterior aspect of the spinal cord. The wave form of the response differed according to the level of the stimulating epidural electrodes. Cervical stimulation elicited an SEP very similar to that produced by stimulation of upper extremity nerves, e.g., bilateral median nerve SEP, but with a shorter latency. Epidural stimulation of the lower thoracic cord elicited an SEP similar to that produced by stimulation of lower extremity nerves. The results of upper thoracic stimulation appeared as a mixed upper and lower extremity type of SEP. The overall amplitudes of SEPs elicited by the epidural stimulation were higher than SEPs elicited by peripheral nerve stimulation. In 4 patients the CV along the spinal cord was calculated from the difference in latencies of the cortical responses to stimulation at two different vertebral levels. The CVs were in the range of 45-65 m/sec. The method was shown to be promising for future study of spinal cord dysfunctions.     

Short latency somatosensory evoked potentials from radial, median and ulnar nerve stimulation in man

Short latency somatosensory evoked potentials (SEPs) were elicited by stimulation at the wrist of median, radial, and ulnar nerves, singly or in combination, using normal subjects. Amplitude of P10 was strikingly lower with radial stimulation than with median stimulation, while ulnar-derived P10 was intermediate in amplitude. This difference probably reflects the antidromic firing of motor fibers contained in median nerves as compared with the superficial branch of radial nerve, which is entirely sensory. Beyond P10, there appear to be no significant differences between median, radial and ulnar-derived SEPs. With simultaneous stimulation of several nerves within one arm, larger potentials were sometimes achieved but with poorer definition of P12 and P14. The clinical utility of radial, ulnar, and median stimulation for localizing peripheral lesions derives from the distinct anatomical pathways of the stimulated fibers through the brachial plexus and from the separable motor and sensory components of P10. SEP is less invasive than EMG; this fact, plus its freedom from sampling error, make it potentially more suitable than conventional EMG for sequentially following a patient's clinical course.     

The investigation of traumatic lesions of the brachial plexus by electromyography and short latency somatosensory potentials evoked by stimulation of multiple peripheral nerves.

A study of 10 patients with brachial plexus trauma was performed to determine whether the diagnostic accuracy of sensory evoked potentials (SEPs) may be improved by using stimulation of multiple peripheral nerves (median, radial, musculocutaneous and ulnar). In addition, the relative advantages of SEPs and peripheral electrophysiological studies were considered. SEP patterns following most common brachial plexus lesions were predictable. Injuries to the upper trunk affected the musculocutaneous and radial SEPs predominantly. Lower trunk or medial cord lesions primarily affected ulnar SEPs. Diffuse brachial plexus lesions affected SEPs from all stimulation sites. In the majority of cases, the necessary information was obtainable from conventional EMG: however, for lesions involving the upper segments only, SEP techniques were more useful. It is suggested that selective SEPs from appropriate peripheral nerves when interpreted in combination with conventional EMG may add useful additional information.     

A comparison of forepaw and vibrissae somatosensory cortical evoked potentials in the rat

Somatosensory evoked potentials were elicited in anesthetized rats by electrical stimulation of the forepaw (F-SEP) or the vibrissae (V-SEP) and were compared in order to study which of these is a more valid animal model for studying the physiology and pathophysiology of somatosensory evoked potentials (SEPs) that are often recorded in man in a clinical setting. Intensity and rate functions were measured for the two potentials. The V-SEPs had larger amplitudes than the F-SEPs at high stimulus intensity and low stimulus rate. Furthermore, the ratios of the maximal amplitude of the F-SEP to that of the V-SEP (0.66) and of the areas under the curves of the two responses (0.75) reflected the smaller representation of the forepaw in the primary somatosensory cortex of the rat, compared to the vibrissae (ratio of cortical areas about 0.79). The differences should be taken into account when using median nerve SEP in the rat as a model of the human SEP. Study of V-SEPs in rat may provide insight into trigeminal nerve SEPs in man, which are also occasionally used for neurological evaluation.     

Prognostic value of early somatosensory evoked potentials in the presence of traumatic lesions of the spinal cord

Cortical and cervical early somatosensory evoked potentials (SEPs) were recorded after stimulation of median, ulnar, tibial or common peroneus nerves in 23 patients with traumatic paraplegia or quadriplegia. The clinical progress of these patients was followed for more than 18 months. The gradual clinical recovery or its absence was compared with neurophysiological data. In complete spinal injury, the absence of SEPs in response to stimulation of a nerve entering the cord below the level of injury demonstrates the high degree of spinal cord damage. There was a good correlation between the presence of SEP, even of low amplitude or long latency, and a favorable prognosis. The return of the SEP could herald clinical recovery of posterior column function. From a practical point of view, these preliminary data suggest that the investigation of early SEPs must begin with a very short delay after injury and be continued for approximately 6 months so that the irreversibility of the lesion can be clearly established.     

Middle-latency somatosensory evoked potentials after stimulation of the radial and median nerves: component structure and scalp topography.

Somatosensory evoked potentials (SEPs) after radial nerve stimulation are studied less frequently than those after median nerve stimulation. Therefore, little is known about their component structure and scalp topography. We investigated radial nerve SEPs after electrical stimulation at the left wrist. For comparison, the median nerve was also stimulated at the wrist. SEPs were recorded with 15 scalp electrodes (bandpass 0.5-200 Hz) in 27 healthy subjects. The waveform of the radial nerve SEP at a contralateral parietal lead was comparable to that of the median nerve SEP, consisting of P14, N20, P30, and N60. In spite of comparable stimulus intensities, SEP amplitudes were smaller after radial than after median nerve stimulation. Significant latency differences were found only for N20 (earlier for median nerve) and P30 (earlier for radial nerve). The duration of the primary complex N20-P30 thus was significantly shorter for the radial nerve. Whereas N20 and P30 were present with either earlobe or frontal reference, N60 had a prerolandic maximum and was best recorded with a bipolar transverse derivation. In addition, another middle-latency negativity (N110) was found near the secondary somatosensory cortex, which had previously been described only for radial nerve stimulation. In standard SEP derivations, the N110 is riding on the ascending limb of the vertex negativity. It could best be recorded in low temporal leads versus a midline reference. The scalp topographies of P30, N60, and N110 were similar for radial and median nerve stimulation.     

Sensory deficits of a nerve root lesion can be objectively documented by somatosensory evoked potentials elicited by painful infrared laser stimulations: a case study.

Somatosensory evoked potentials (SEPs) in response to painful laser stimuli were measured in a patient with a unilateral sensory deficit due to radiculopathy at cervical levels C7 and C8. Laser evoked potentials (LEPs) were compared with SEPs using standard electrical stimulation of median and ulnar nerves at the wrist and mechanical stimulation of the fingertips by means of a mechanical stimulator. Early and late ulnar and median nerve SEPs were normal. Mechanical stimulation resulted in w shaped early SEPs from all five fingertips with some degree of abnormality at the fourth and fifth digits of the affected hand. Late LEPs were completely absent for stimulations at affected dermatomes and normal in the unaffected control dermatomes. The border between skin areas with normal or absent LEPs was very sharp and fitted the dermatomes of intact C6 and damaged C7 and C8 nerve roots. It is suggested that pain dermatomes are narrower than tactile dermatomes because thin fibres of the nociceptive system, activated by laser stimuli, probably do not overlap between adjacent spinal segments to the same extent as thick fibres of the mechanoreceptive system, activated by standard electrical or mechanical stimulation.     

Neuromuscular consequences of prolonged hunger strike: an electrophysiological study.

OBJECTIVES: The purpose of this study was to determine the electrophysiological consequences of neuromuscular and central nervous system involvement in a group of patients presented with the neurological complications of a long-term hunger strike (HS). METHODS: Motor and sensory nerve conduction (NCV), F wave, somatosensory evoked potential (SEP) and motor evoked potential (MEP) studies were performed in 12 male and 3 female patients (mean age: 29.4) following HS. RESULTS: All patients whose weight loss was 11-31 (mean: 22.8) kg after 69-day HS, had neurological findings consistent with Wernicke's encephalopathy or Wernicke-Korsakoff syndrome. Abnormally prolonged latency and/or low amplitude sensory nerve action potentials were found in 7 patients. The amplitudes of compound muscle action potentials were significantly reduced in ulnar, median and tibial motor NCV studies as compared to the controls. F waves elicited by median nerve stimulation at wrist and muscle responses evoked by cervical and lumbar magnetic stimulation had significantly prolonged latencies. MEPs recorded from the lower extremities showed a slight prolongation in central conduction times. The cortical response latencies were prolonged in tibial SEPs. CONCLUSIONS: The most prominent finding in this patient group was the low amplitude of CMAPs elicited in motor NCV studies which was concluded to be resulted from the reversible muscular changes. The other electrophysiological findings suggested that peripheral nerves and long central nervous system pathways were also mildly involved.     

Nerve, spinal cord and brain somatosensory evoked responses: a comparative study during electrical and magnetic peripheral nerve stimulation.

Somatosensory evoked potentials (SEPs) and compound nerve action potentials (cNAPs) have been recorded in 15 subjects during electrical and magnetic nerve stimulation. Peripheral records were gathered at Erb's point and on nerve trunks at the elbow during median and ulnar nerve stimulation at the wrist. Erb responses to electrical stimulation were larger in amplitude and shorter in duration than the magnetic ones when 'electrical' and 'magnetic' compound muscle action potentials (cMAPs) of comparable amplitudes were elicited. SEPs were recorded respectively at Cv7 and on the somatosensory scalp areas contra- and ipsilateral to the stimulated side. SEPs showed a statistically significant difference in amplitude only for the brachial plexus response and for the 'cortical' N20-P25 complex; differences were not found between the magnetic and electrical central conduction times (CCTs) or for the peripheral nerve response latencies. Magnetic stimulation preferentially excited the motor and proprioceptive fibres when the nerve trunks were stimulated at motor threshold intensities.     

Short-latency somatosensory-evoked potentials from radial, median, ulnar, and peroneal nerve stimulation in the assessment of cervical spondylosis. Comparison with conventional electromyography

A study of data on 30 patients with cervical spondylosis was carried out to determine whether short-latency somatosensory-evoked responses (SEPs) to median, ulnar, radial, and peroneal nerve stimulation provided additional information to that obtained by electromyography (EMG), late responses, and peripheral conduction studies. Peripheral studies, EMG results and SEPs were within normal limits in ten patients with pain, but without objective neurological deficit. By contrast, of ten patients who had objective signs of root compression, conventional EMG results were normal in nine, but abnormalities of the SEPs from radial nerve stimulation were obtained in only five patients, and were normal from ulnar and median nerve stimulation. In ten patients with clinical features of myelopathy, seven had abnormal median SEPs and all had abnormal peroneal SEPs, whereas EMG results were abnormal in only five patients. It is suggested that SEPs and EMG are both of limited use in patients with only symptoms of root compression. In patients with signs of root compression, EMG is the most sensitive procedure; however, some additional information can be obtained from superficial radial SEPs. In patients with cervical myelopathy, SEP was the most useful procedure, especially when upper and lower limbs were studied.     

The value of ulnar somatosensory evoked potentials (SEPs) in cervical myelopathy

In 57 patients with clinical signs and surgical documentation of compressive myelopathy, ulnar nerve somatosensory evoked potentials (SEPs) were more sensitive (with 74% abnormal) than either median or tibial nerve SEPs. The most frequent abnormalities were reduced or absent neck evoked responses and prolonged central conduction time. All subjects who had an SEP abnormality were identified by combined tibial and ulnar SEPs. Median nerve SEP added no additional information. Normal ulnar and tibial nerve SEPs were also able to exclude major cord damage in patients with cervical radiculopathy but little evidence of myelopathy.     

Segmental dysfunction of the cervical cord revealed by abnormalities of the spinal N13 potential in cervical spondylotic myelopathy.

We studied somatosensory potentials (SEPs) evoked by stimulation of radial, median, and ulnar nerves in 11 patients with MRI evidence of cervical spondylosis. All patients presented with progressive spastic paraparesis that was either isolated or associated with lower motor neuron signs in the upper limbs, with preserved joint, touch, pain, and temperature sensations in the four limbs. In all patients, scalp SEPs reflecting the activity of the dorsal column system up to the parietal cortex were normal while segmental cervical cord dysfunction was manifested by an abnormal spinal N13 potential in 95% of radial, 90% of median, and 54% of ulnar nerve SEPs. These subclinical abnormalities of the spinal N13 SEP probably result from reduced blood supply due to compression of the anterior spinal artery in patients with cervical spondylotic myelopathy.     

Thalamo-cortical processing of near-threshold somatosensory stimuli in humans

Somatosensory stimuli elicit complex cortical responses that are discernible as somatosensory evoked potentials (SEPs) in scalp electroencephalographic recordings. Whereas earlier SEP components, occurring up to 100 ms after stimulus delivery, have been labeled 'preconscious', later responses have been associated with stimulus awareness. To date, how far these processes are primarily cortical or comprise additional subcortical operations remains open. Therefore, we recorded thalamic and scalp SEPs evoked by perceived as well as unperceived median nerve stimulation in neurosurgical patients with electrodes implanted into the ventral intermediate nucleus of the thalamus for deep brain stimulation. At stimulation intensities below perceptual threshold, only thalamic SEP components appeared consistently during the first 75 ms after stimulus delivery. Stimulation that was perceived by the patients elicited cortical as well as thalamic SEPs that lasted longer than 75 ms. These results indicate that the thalamus remains active after the primary propagation of a sensory signal to the cortex, and suggest that the transition from elementary to higher-order somatosensory processing is based on thalamo-cortical interactions.     

Somatosensory evoked potentials after multisegmental upper limb stimulation in diagnosis of cervical spondylotic myelopathy.

Radial, median, and ulnar nerve somatosensory evoked potentials (SEPs) were recorded, with non-cephalic reference montage, in 38 patients with clinical signs of cervical myelopathy and MRI evidence of spondylotic compression of the cervical cord. Upper limb SEPs are useful in spondylotic myelopathy because SEPs were abnormal in all patients for at least one of the stimulated nerves and SEP abnormalities were bilateral in all patients but one. Reduction of the amplitude of the N13 potential indicating a segmental dysfunction of the cervical cord was the most frequent abnormality; it occurred in 93.4%, 84.2%, and 64.5% of radial, median, and ulnar nerve SEPs respectively. A second finding was that the P14 far-field potential was more sensitive than the cortical N20 potential to slowing of conduction in the dorsal column fibres. The high percentage of N13 abnormalities in the radial and median rather than in the ulnar nerve SEPs correlated well with the radiological compression level, mainly involving the C5-C6 vertebral segments. Therefore the recording of the N13 response is a reliable diagnostic tool in patients with cervical spondylotic myelopathy and P14 abnormalities, though less frequent, can be useful in assessing subclinical dorsal column dysfunction.     

Sympathetic skin response in hemispheric lesions.

We recorded the sympathetic skin response (SSR) from electrical nerve stimulation in 16 patients with cerebrovascular accident (CVA). Location and nature of the lesion were documented by computerized tomography (CT). Median (pre-rolandic and parietal) somatosensory evoked potentials (SEP) were also recorded. SSR was absent bilaterally in eight hemiplegics after stimulation of the plegic side and present bilaterally after stimulation of the normal side in the first weeks after CVA. Parietal and pre-rolandic SEPs were absent in the affected hemispheres. SSR was present bilaterally after stimulation of each side in the remaining CVA cases with reduced amplitude SEPs. Absence of the electrically evoked SSR in hemispheric lesions may be due to involvement of central afferent pathways or temporary suppression of suprasegmental excitatory influences.     

Source generators of the early somatosensory evoked potentials to tibial nerve stimulation: an intracerebral and scalp recording study.

OBJECTIVE: To investigate the location of the cerebral generators of the early scalp somatosensory evoked potentials (SEPs) after tibial nerve stimulation. METHODS: Tibial nerve SEPs were recorded in 15 patients, suffering from Parkinson's disease, who underwent implantation of intracerebral (IC) electrodes in the subthalamic nucleus, in the globus pallidum or in the thalamic ventralis intermediate nucleus. SEPs were recorded both from the scalp surface and from the IC leads. RESULTS: The lemniscal P30 response was recorded by all the electrodes. The IC waveforms included a negative N40IC response, followed by a positive (P50IC) and a negative (N60IC) potential. The N40IC, the P50IC and the N60IC potentials did not differ in latency from the P40, the N50 and the P60 responses recorded by the Cz electrode. In 6 patients, in which SEPs were recorded also during the voluntary movement of the stimulated foot (active gating), an amplitude reduction of the SEP components following the P30 potential was observed during movement at the vertex and in the IC traces. Instead, in the contralateral temporal traces the SEP components (N40temp and P50temp) were not modified by active gating, and in the ipsilateral parietal traces only the positive potentials at about 60ms of latency was decreased. CONCLUSIONS: Two differently oriented generators are active in the contralateral hemisphere at both 40 and 50ms of latency after tibial nerve stimulation. One source is oriented perpendicularly to the mesial hemispheric surface and generates the potentials recorded by the contralateral temporal and the ipsilateral parietal leads; the other dipolar source is radial to the hemispheric convexity, and generates the potentials at the vertex and those recorded by the IC electrodes.     

Early-latency somatosensory evoked potentials elicited by electrical acupuncture after needling acupoint LI-4.

The stimulating methods of prior studies on somatosensory evoked potentials (SEPs) elicited by acupoint stimulus had involved surface electrodes, while the clinical practice of acupuncture is mostly performed by inserting the acupuncture needle inside the body. Clinical observations show that there are often some special sensations when LI-4 is needled. To observe if the SEPs produced by acupoint acupuncture had a distinguishing property, we studied the SEPs elicited by electrical acupuncture after the acupuncture needle was inserted into LI-4 and its control point, and then mapped them with the 128-channel Electric Brain Signal Image system. We also compared this to SEPs by median nerve stimuli. Results showed that the most interesting finding was the marked differences of N1-P1 and N2-P2 amplitude between SEPs at LI-4 (SEP-LI) and its control point (SEP-CP), which were in the opposite direction. Marked differences were also found between latencies and amplitudes of the SEPs elicited by acupuncture and by median nerve stimulus (SEP-M). The differences between SEP-LI and SEP-CP might be due to the additional effects of the activation of nerve endings and muscle spindles in LI-4 to the SEPs formed by the activation of superficial and deep radial nerves during electrical acupuncture. The differences between SEPs to acupoint and median nerve stimuli might be mainly due to the different distances from the stimulated regions to the cerebral cortex, the diversity and the number of activated fibers.     

Retrograde effects of digital nerve severance on somatosensory evoked potentials in man

To determine if retrograde conduction changes might occur long after injury of the most distal peripheral nerves, short-latency somatosensory evoked potentials (SEPs) to median or ulnar nerve stimulation at the wrist were studied in 10 subjects who had sustained traumatic digit amputation 4 months to 15 years previously. SEPs were recorded from Erb's point (N9), the cervical region (N13), and the contralateral scalp hand area (N20). While N9 latency was slightly delayed or not affected, the amplitude was either markedly reduced or undetectable. For N13 and N20 components, latency prolongation and amplitude reduction were mild to moderate, but the central conduction time (N13–N20) remained normal. The present data indicate that even the most distal nerve injury may have profound long-term retrograde effects on parental nerve function which are presumed mainly due to an axonopathy. © 1994 John Wiley & Sons, Inc.     

Influence of the ipsilateral hemisphere upon somatosensory evoked potential in cats

Somatosensory evoked potentials (SEPs) were recorded on the skull corresponding to the (left) SI area by (right) superficial radial nerve stimulation. Amongst the various components of the SEP, special attention was directed to the negative component (N15) with a latency of approximately 15 ms. Changes in this potential followed by conditioning stimulation of the ipsilateral (right) hemisphere were observed and the following results were obtained: (i) when conditioning stimuli were applied to the contralateral (left) superficial radial nerve, the ipsilateral (right) thalamic VPL nucleus and the ispilateral (right) sensory cortex, the amplitude of N15 decreased to 65-80% of the control level at C-T intervals less than 100 ms and (ii) following functional elimination of the unilateral sensory cortex by KCI application, the amplitude of N15 recorded at the opposite side significantly increased. In this condition, the inhibitory effects of the ipsilateral thalamus and contralateral peripheral nerve disappeared. From these observations, ipsilateral homologous cortex may well have an inhibitory influence upon the near field potential (N15) of the SEP.     

Changes in median nerve somatosensory transmission and motor output following transient deafferentation of the radial nerve in humans.

OBJECTIVE: To determine if transient anaesthetic deafferentation of the radial nerve would lead to alterations in processing of early somatosensory evoked potentials (SEPs) from the median nerve or alter cortico-motor output to the median nerve innervated abductor pollicis brevis (APB) muscle. METHODS: Spinal, brainstem, and cortical SEPs to median nerve stimulation were recorded before, during and after ipsilateral radial nerve block with local anaesthesia. Motor evoked potentials (MEPs) and motor cortex output maps were recorded from the APB muscle. RESULTS: There were no significant changes to most early SEP peaks. The N30 peak, however, showed a significant increase in amplitude, which remained elevated throughout the anaesthetic period, returning to baseline once the anaesthetic had completely worn off. MEP amplitude of the median nerve innervated APB muscle was significantly decreased during the radial nerve blockade. There was also a significant alteration in the APB optimal site location, and a small but significant decrease in the silent period during the radial nerve blockade. CONCLUSIONS: Transient anaesthetic deafferentation of the radial nerve at the elbow leads to a rapid modulation of cortical processing of median nerve input and output. These changes suggest an overall decrease in motor cortex output to a median nerve innervated muscle not affected by the radial nerve block, occurring concomitantly with an increased amplitude of the median nerve generated N30 SEP peak, thought to represent processing in the supplementary motor area (SMA). Independent subcortical connections to the SMA are thought to contribute to the N30 response observed in this study. Unmasking of pre-existing but latent cortico-cortical and/or thalamo-cortical connections may be the mechanism underlying the cortical SEP increases observed following radial nerve deafferentation. SIGNIFICANCE: Transient deafferentation of the radial nerve, which supplies wrist and hand extensor muscles, has been shown to alter sensory processing from and motor output to the median nerve innervated thenar muscles.