- Email: [email protected]
Changes in somatosensory evoked responses by repetition of the median nerve stimulation
Clinical Neurophysiology 114 (2003) 2251–2257 www.elsevier.com/locate/clinph
Changes in somatosensory evoked responses by repetition of the median nerve stimulation Minoru Hoshiyamaa,b,*, Ryusuke Kakigia a
Department of Integrative Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444-8585, Japan b Department of Health Sciences, Faculty of Medicine, Nagoya University, Nagoya 461-8673, Japan Accepted 30 July 2003
Abstract Objective: We investigate the synaptic factor for the recovery function of evoked responses using a repetitive stimulation technique. Methods: Somatosensory evoked cortical magnetic field (SEF) was recorded following stimulation of the median nerve using single to 6-train stimulation in 8 healthy subjects. The SEF responses after each stimulus in the train stimulation were extracted by subtraction of the waveforms. Results: An attenuation of the SEF components was recognized after the second of the stimuli, but there was no significant attenuation with the third or later stimulations. The root mean square (RMS) of the 1M (peak latency at 20 ms after stimulation) and 4M (70 ms) components were smaller than that of the single stimulation during the train stimulation, while the 2M (30 ms) and 3M (45 ms) components were not attenuated, but the 3M was facilitated at the fourth to sixth stimulation. Conclusion: The synaptic factor was not responsible for the attenuation of the SEF components during repetitive stimulation in healthy subjects. The SEF change disclosed a functional difference among the SEF components during the train stimulation, especially among the later components. q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Somatosensory evoked potential; Somatosensory evoked cortical magnetic field; Repetitive stimulation; Magnetoencephalography; Human; Synaptic depletion
1. Introduction In the study of the recovery function of somatosensory evoked potential (SEP) and cortical magnetic field (SEF), the amplitude of the components decreases as the interstimulus interval (ISI) of the paired stimulation is shortened (Allison, 1962; Meyer-Hardting et al., 1983; Emori et al., 1991; Hoshiyama and Kakigi, 2001, 2002). Previously, we reported that each SEP/SEF component was refractory to a second stimulus after an initial stimulus with a very short ISI (0.5 – 100 ms), and concluded that a structural or functional process of refractoriness existed for high-frequency signals in the primary sensory cortex and there was a difference in refractoriness among the SEP/SEF components (Hoshiyama and Kakigi, 2001, 2002). We speculated, from the results of basic animal studies (Cooley and Dodge, 1966; Fitzhugh, * Corresponding author. Tel.: þ 81-564-55-7779; fax: þ81-564-52-7913. E-mail address: [email protected] (M. Hoshiyama).
1961; Parnas et al., 1976; Waxman, 1972), that one of the factors responsible for the refractoriness to high frequency signals was the morphological structure of the tapering of the axon. However, considering the depletion of the synaptic transmitter following high frequency stimulation, the function of neural conduction including transmission of the signal via synapses might limit the response to the high frequency signals (Stevens and Wang, 1995). Although we considered that the number of synapses was not the major factor for the recovery function of the SEP/SEF components (Hoshiyama and Kakigi, 2001, 2002), there remained the possibility that the difference of synaptic type caused the difference in the recovery function, or the refractoriness to the high frequency signals. We defined the depletion of the synaptic transmitter following high frequency stimulation as ‘synaptic factor’ in the present study. Therefore, the objective of the present study was to investigate the synaptic factor for the recovery function of the responses evoked using a repetitive stimulation
1388-2457/$30.00 q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S1388-2457(03)00285-2
2252
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
technique. The theoretical basis of this study was similar to that of the repetitive stimulation test to investigate the function of the neuromuscular junction, except that we used a multi-channel magnetoencephalography (MEG) system. In addition to the high reliability of dipole estimation, the stimulus artifacts are much smaller in MEG than electroencephalogram recordings, and MEG with axial gradiometers has the advantage of recording cortical responses with a high signal to noise ratio. We used these advantages of MEG for the SEF recording following repetitive stimulation.
2. Methods 2.1. Subjects Eight male volunteers in our department (aged 26 – 39 years) were examined. Informed consent to participate in the study, which was first approved by the Ethical Committee of the National Institute for Physiological Sciences, Okazaki, Japan, was obtained from all participants prior to the study. The subjects were not told about the prospective results of the study prior to the recording. In the present study, the initial 4 components of SEF were investigated (1M, 2M, 3M and 4M, see Section 3). Since the interindividual difference of the later components (3M and 4M) was considerably larger than that of the first two components (1M and 2M), only subjects in whom the 4 components could be identified in a pilot study, participated in the study.
and the left parietal area were mostly covered by the position. 2.3. Stimulus conditions The stimulation was a single to 6-train variety with 3 different ISIs, i.e. 10, 20 and 30 ms. Thus, we used 16 sets of stimuli, which comprised single stimulation, double to 6-train stimulation with a 10 ms ISI, double to 6-train stimulation with a 20 ms ISI and double to 6-train stimulation with a 30 ms ISI. The number of stimuli in the train was expressed with a capital roman numeral followed by the ISI, e.g. V-10 indicated a 5 train stimulation with a 10 ms ISI. Those 16 conditions were controlled by a computer system (Power-1401, CED, Cambridge, UK). The interval between the offset of the last stimulation of a train and the onset of the following train stimulation was 1s (Fig. 1). Each single or train stimulus was randomly delivered to the right median nerve at the wrist using a saddle type electrode. The electrical stimulus was a constant current square-wave pulse of 0.2 ms duration. The intensity of each stimulus was adjusted to produce a slight twitch of the thumb, which was sufficient to produce a SEF of saturated amplitude to a single stimulus. Average intensity was 4.6 mA (range 3.7 –5.3 mA). The duration of the recording epoch was 300 ms after the onset of the train stimulation. The bandpass filter was 1 – 200 Hz with sampling of 2048 Hz.
2.2. SEF recording We focused on the major short- and middle-latency cortical SEF components, i.e. the 4 components appearing approximately 20 ms (1M), 30 ms (2M), 50 ms (3M) and 70 ms (4M) after the stimulation, which were recorded around the central sulcus in the hemisphere contralateral to the stimulated side. The magnetic fields were measured with 37-channel axial-type gradiometers (Magnes, Biomagnetic Technologies Inc., San Diego, CA) in a magnetically shielded room. The detection coils of the system were arranged in a uniformly distributed array in concentric circles over a spherically concave surface. Thus, all of the sensor coils were equally sensitive to the brain’s weak magnetic signals. The device was 144 mm in diameter and its radius was 122 mm. The outer coils were 72.58 apart. Each coil was 20 mm in diameter and the distance between the centers of the coils was 22 mm. Each coil was connected to a superconducting quantum interference device. The measurement matrix was centered at around C3 of the International 10 – 20 system in each subject. The left primary (SI) and secondary (SII) somatosensory cortices
Fig. 1. Experimental procedure and the subtraction method for the SEF analysis. (A) Single to 6-train stimulations with ISIs of 10, 20 and 30 ms were randomly delivered to the median nerve. The number of stimuli in the train was expressed with a capital roman numeral followed by the ISI, e.g. III-30 indicated a 3 train stimulation with an ISI of 30 ms. The interval between the last stimulation and the onset of the first of the following stimulations was 1 s. (B) The SEF waveform evoked by the last stimulation of a train was obtained by subtracting the waveform of a less numerous train stimulation from the waveform, e.g. the SEF waveform evoked by the fifth stimulation (v) was obtained by subtracting the 4-train stimulation (IV) from the 5-train stimulation (V).
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257 Table 1 The peak latency and the RMS values of the SEF components following single stimulation SEF components 1M
2M
3M
2253
with a digital marking in each stimulus train condition. One session took approximately 50 min, and, a short rest was given every 10 min during the recording. The recording was repeated on another day to test the reproducibility of the responses.
4M
Peak latency 18.9 ^ 0.9 28.9 ^ 1.7 45.5 ^ 3.2 72.6 ^ 7.2 (mean ^ SD, ms) RMS 85.5 ^ 30.6 68.1 ^ 38.7 49.0 ^ 19.7 104.4 ^ 32.6 (mean ^ SD, fT)
We setup the recording equipment so as to collect all of the data in one session to minimize the intra-individual variation in the MEG signal. One recording session comprised 200 epochs for each stimulus condition. Two hundred epochs were collected separately for each condition
2.4. Data analysis The MEG signals were collected and stored on a magneto-optical disk, and analyzed for later processing by an off-line system. The epochs of each stimulus condition were collected and averaged separately. The SEF waveform evoked by the last stimulation of a train was obtained by subtracting the waveform of a less numerous train stimulation from the waveform, and the waveform was expressed with a small roman numeral, e.g. the SEF
Fig. 2. The SEF waveform following each train stimulation in one subject. The number of stimuli in the train was expressed as a capital roman numeral, and the arrow indicates the stimulation.
2254
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
waveform evoked by the fifth stimulation in the V-10 was obtained by subtracting IV-10 from V-10, and the waveform was described as v-10. The peak latency of first 4 components of SEF and the root-mean-square (RMS) value at the peak of the components was calculated from the 37-channel waveform. Those values were assessed to analyze each stimulus condition. The peak latency and the RMS values were expressed as a ratio, which was obtained by dividing each value by the value of the SEF following single stimulation (Table 1). The two-way (number of repetition and ISI sequence) repeated measures analysis of variance (ANOVA) was used for statistical analysis of the latency or RMS value. Bonferroni – Dunn’s correction was used for multiple comparisons, when necessary. 3. Results Fig. 2 shows the raw averaged SEF waveforms following each stimulus train recorded from one subject. In the waveforms following the single stimulation, 4 SEF components were identified. The components were termed 1M, 2M, 3M and 4M, and the peak latency and the RMS value at the peak are shown in Table 1. The stimulus artifact was not considerable, which is one of the major advantages of SEF recordings, and subtracted
waveforms were successfully obtained (Fig. 3). The RMS values of the 1M and 4M components attenuated with the repetition of the stimulation in all ISI sequences (P , 0:01), and the attenuation was greater in the 10 ms ISI sequence than in the 20 and 30 ms ISI sequences (P , 0:02) (Fig. 4). However, there was no significant change among the number of repetitions. For the 2M and 3M components, there was evidently a different change in the RMS values. The RMS values of the 2M component did not decline, but tended to increase at iii-10 and ii-20, although the difference was not significant (P ¼ 0:051, ANOVA with Bonferroni – Dunn’s correction) (Fig. 4). For the 3M component, the RMS values increased in the 20 and 30 ms ISI sessions with the number of repetitions (P , 0:02, ANOVA), and the increase at ii-10 was significant (P , 0:02, ANOVA with Bonferroni – Dunn’s correction). The latency was prolonged in general (Fig. 5). The prolongation of the 1M component was significantly greater in the 10 ms ISI sequence than in the 20 and 30 ms ISI sequences (P , 0:02, ANOVA). However, for the 2M, 3M and 4M components, there was no significant difference in the prolongation among the sequences or among the number of stimulus repetitions. The latency ratio of the 3M component was significantly greater than that of 1M in the 20 and 30 ms sequences (P , 0:02, ANOVA).
Fig. 3. The SEF waveform evoked by the last stimulation of a train obtained by the method shown in Fig. 1. The small roman numerals (i –vi) indicate the number of the last stimulus in the train, e.g. the SEF waveform evoked by the last (fourth) stimulation in IV-10 condition is described as iv-10. The subject was the same as in Fig. 2.
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
Fig. 4. The RMS changes of the SEF components during the stimulus repetition. Each value was expressed as a ratio to that of the SEF waveform following single stimulation. The small roman numerals (ii–vi) indicate the stimulus number in the train. The 1M and 4M components were attenuated during the train stimulation, while the 2M and 3M were not. There was no significant decrease in the RMS values in the stimulus train, but for the 3M component, the RMS values rose in the 20 and 30 ms ISI sessions as the number of repetitions increased (P , 0:02, ANOVA), and the RMS values of 3M at the fourth (iv) to sixth (vi) stimulations with ISIs of 20 and 30 ms were greater than those of the second (ii) and third (iii) stimulations (*P , 0:02, ANOVA, Bonferroni–Dunn’s correction).
4. Discussion In the present study, we confirmed that there was an attenuation of the SEF components in terms of amplitude following the second of a train of stimuli but no waning or waxing of the response was recognized between the third and sixth stimuli, and that the pattern of attenuation differed among the SEF components. This is the first report of the change in the evoked response during repetitive stimulation with a very short ISI in humans, although the recovery function of the evoked responses (Allison, 1962; Shagass and Shwartz, 1964; Wiederholt, 1978; Meyer-Hardting et al., 1983; Emori et al., 1991; Hoshiyama and Kakigi, 2001, 2002) and the decrease in SEP during repetitive stimulation with an ISI of more than 500 ms (Angel et al., 1985) were previously reported. Previous studies reported an attenuation of SEP components evoked by stimulation with a relatively high stimulus rate of up to 30 Hz (Pratt et al., 1980; Abbruzzese et al., 1990; Delberghe et al., 1990; Fujii et al., 1994), and that the majority of the loss occurred after the second
2255
stimulation in a repetitive series with an ISI of between 500 and 800 ms (Angel et al., 1985). Saito et al. (1992) and Fujii et al. reported that the mechanism of the attenuation in the recovery function of SEF was partially due to interference between the electrical stimulation and the secondary afferent input from the muscle contraction. Since the muscle contracted during the train stimulation also in the present study and we could not know whether the afferent signal from the muscle was stable or not during the train stimulation, we could not exclude the effects of muscle contraction on the SEF changes, especially on the SEF change following the second stimulation. However, since the decrement of the 1M and 2M components were not affected by the stimulus repetition during the second to the sixth stimulation, we considered that we could focused on another point. We considered that the functional refractoriness of the neuron might partially relate to the synaptic ability to respond to high frequency stimulation, and that the SEP change during repetitive stimulation with a very short ISI would give the information on the synaptic factor of the central nervous system. For example, when the half of the synapses did not respond to the second stimulation, the half of the rest of the synapses and the synapse, which did not respond to the second stimulation, could respond to the third one. The total amount of synapses responded to the third stimulation was 75% of the synapses. Similarly, if the synaptic factor mainly contributed to the attenuation of an SEF component during repetitive stimulation, the response might change in amplitude between the second and later components. The decrement of the 1M component was approximately 50% following the second stimulation at an ISI between 10 and 30 ms, and there was no significant change of the value at the following stimulations. Therefore, the synaptic factor was unlikely to be a major factor in the SEF attenuation during the double and repetitive stimulation, at least at an ISI of 10 –30 ms in healthy subjects, although we could not exclude the partial contribution of the synaptic factor. In the in vitro study using pyramidal neurons in slices of the hypocampal region, a single synapse seemed to require up to 20 ms to refill the transmitter (Stevens and Wang, 1995). The discrepancy between their results and our’s might be explained by that the depression of the transmission for a short ISI was not profound in a group of neurons, as Stevens and Wang (1995) described. However, at an ISI of less than 10 ms, the transmission failure at synapses might occur and cause the attenuation of the SEP/SEF components. In our previous study, we speculated that tapering axons might contribute to the SEF attenuation in the recovery function (Hoshiyama and Kakigi, 2001). We considered a possibility that the morphological structure of axon might responsible for the change of attenuation of 1M in the present study. If the morphological structure was responsible for the SEF attenuation at high frequency stimulation, the attenuation could be stable during repetitive stimulation.
2256
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
Fig. 5. The latency change during the stimulus repetition. Each value was expressed as a ratio to that of the SEF waveform following single stimulation. The small roman numerals (ii–vi) indicate the stimulus number in the train. The latency was prolonged in general, and the prolongation was significantly greater for the 10 ms than 20 or 30 ms ISI sequence (P , 0:02, ANOVA), and the prolongation of the 3M latency was significantly greater than that of 1M in the 20 and 30 ms sequences (P , 0:02, ANOVA). There was no significant difference in prolongation among the sequences or among the number of stimulus repetitions for 2M, 3M and 4M.
The pattern of attenuation for 1M was evidently different from that for 2M. The 1M components showed a decrease during repetitive stimulation, but the 2M component did not show any significant attenuation. These results were essentially consistent with the results of the previous study (Hoshiyama and Kakigi, 2001). Differences in the generator or structure of the neurons, including axonal structure, were speculated as the reason for the result (Hoshiyama and Kakigi, 2001, 2002), which might cause the functional difference in the response to ‘secondary afferent signals’ (Fujii et al., 1994). With regard to the later responses, the 3M component did not show any significant attenuation during the repetitive stimulation, but it was strengthened at the fourth to sixth stimuli. For the 1M and 2M components, the major generator has been estimated to exist around the primary somatosensory cortex (SI) (Hari et al., 1984; Wood et al., 1985; Kakigi, 1994; Hoshiyama and Kakigi, 2001), but, at the latency of 3M, the activity in the secondary somatosensory cortex (SII) (Karhu and Tesche, 1999; Wegner et al., 2000) and posterior parietal cortex started to overlap (Forss et al., 1994). The SII response was sustained during the repetitive stimulation, while the response in SI was sharp and transient (Forss et al., 2001). The facilitation of 3M by stimulus repetition might be due to the difference in the timing and strength of the inhibition followed by
the excitation (Forss et al., 2001), and, in the present study, sustained muscle contraction during the train stimulation might contributed to the change of the component. On the other hand, the 4M component showed an attenuation in amplitude during the stimulus repetition. The 3M and 4M components could contain the SII activity, but the present results suggested that the function of the 4M component was different from that of the 3M component. Concerning the similarity of the change between the 1M and 4M components during the train stimulation, we could not confirm the mechanism from the present results, although one possibility would be that the 4M response was the secondary response from the 1M signal. The latency of the components was prolonged in general during the repetitive stimulation. The results for the 1M and 2M components were also consistent with the previous results (Hoshiyama and Kakigi, 2001). A prolongation of the latency seemed to be evident for the 3M component. Similarly to the change in amplitude, a functional difference between 3M and the other components was speculated. The jitter of the response might affect more the later components than the early responses, although we are unaware of actual jitter in the SI and SII. Prior to the present study, we tried to record SEP simultaneously with the MEG. However, the electrical artifact from the train stimulation was too large to obtain consistent responses in the subtracted waveform. The correspondence between the SEF and SEP responses should be addressed in a further study using advanced new programs. Abnormal recovery function of SEP has been reported in patients with myotonic dystrophy (Mochizuki et al., 2001), kii amyotrophic lateral sclerosis/parkinsonism-dementia complex (Machii et al., 2003) and subcortical lesions (Ugawa et al., 1996). Those reports disclosed disinhibited cortical excitability in those neurological disorders. The study of the temporal change of SEP/SEF using the method in the present study might give us additional information concerning the cellular and synaptic pathology in those diseases. In conclusion, we reported the change in SEF responses during repetitive stimulation. The attenuation of the amplitude of the SEF component (1M) recognized after the second stimulation did not change with third and later stimulations, but one component (3M) was facilitated by the stimulus repetition. We considered that the synaptic factor was not responsible for the attenuation of the SEF components during repetitive stimulation in healthy subjects. The SEF change disclosed a functional difference among the SEF components during the train stimulation, especially among the later components. References Abbruzzese G, Dall’Agata D, Morena M, Reni L, Trivelli G, Favale E. Selective effects of repetition rate on frontal and parietal somatosensory evoked potentials (SEPs). Electroenceph clin Neurophysiol Suppl 1990; 41:145–8.
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257 Allison T. Recovery function of somatosensory evoked responses in man. Electroenceph clin Neurophysiol 1962;14:331 –43. Angel RW, Quick WM, Boylls CC, Weinrich M, Rodnitzky RL. Decrement of somatosensory evoked potentials during repetitive stimulation. Electroenceph clin Neurophysiol 1985;60:335 –42. Cooley JW, Dodge FA. Digital computer solutions for excitation and propagation of the nerve impulse. Biophys J 1966;6:583–99. Delberghe X, Mavroudakis N, Zegers de Beyl D, Brunko E. The effect of stimulus frequency on post- and pre-central short-latency somatosensory evoked potentials (SEPs). Electroenceph clin Neurophysiol 1990; 77:86–92. Emori T, Yamada T, Seki Y, Yasuhara A, Ando K, Honda Y, Leis AA, Vachatimanont P. Recovery functions of fast frequency potentials in the initial negative wave of median SEP. Electroenceph clin Neurophysiol 1991;78:116 –23. Fitzhugh R. Impulses and physiological states in theoretical models of nerve membrane. Biophys J 1961;1:445–66. Forss N, Hari R, Salmelin R, Ahonen A, Hamalainen M, Kajola M, Knuutila J, Simola J. Activation of the human posterior parietal cortex by median nerve stimulation. Exp Brain Res 1994;99:309–15. Forss N, Narici L, Hari R. Sustained activation of the human SII cortices by stimulus trains. Neuroimage 2001;13:497–501. Fujii M, Yamada T, Aihara M, Kokubun Y, Noguchi Y, Matsubara M, Yeh MH. The effects of stimulus rates upon median, ulnar and radial nerve somatosensory evoked potentials. Electroenceph clin Neurophysiol 1994;92:518 –26. Hari R, Reinikainen K, Kaukoranta E, Hamalainen M, Ilmoniemi R, Penttinen A, Salminen J, Teszner D. Somatosensory evoked cerebral magnetic fields from SI and SII in man. Electroenceph clin Neurophysiol 1984;57:254– 63. Hoshiyama M, Kakigi R. Two evoked responses with different recovery functions in the primary somatosensory cortex in humans. Clin Neurophysiol 2001;112:1334–42. Hoshiyama M, Kakigi R. New concept for the recovery function of shortlatency somatosensory evoked cortical potentials following median nerve stimulation. Clin Neurophysiol 2002;113:535 –41. Kakigi R. Somatosensory evoked magnetic fields following median nerve stimulation. Neurosci Res 1994;20:165–74. Karhu J, Tesche CD. Simultaneous early processing of sensory input in human primary (SI) and secondary (SII) somatosensory cortices. J Neurophysiol 1999;81:2017 –25.
2257
Machii K, Ugawa Y, Kokubo Y, Sasaki R, Kuzuhara S. Somatosensory evoked potential recovery in kii amyotrophic lateral sclerosis/ parkinsonism-dementia complex (kii AlS/PDC). Clin Neurophysiol 2003;114:564–8. Meyer-Hardting E, Wiederholt WC, Budnick B. Recovery function of short-latency components of the human somatosensory evoked potential. Arch Neurol 1983;40:290–3. Mochizuki H, Hanajima R, Kowa H, Motoyoshi Y, Ashida H, Kamakura K, Motoyoshi K, Ugawa Y. Somatosensory evoked potential recovery in myotonic dystrophy. Clin Neurophysiol 2001;112:793–9. Parnas I, Hochstein S, Parnas H. Theoretical analysis of parameters leading to frequency modulation along an inhomogeneous axon. J Neurophysiol 1976;39:909–23. Pratt H, Politoske D, Starr A. Mechanically and electrically evoked somatosensory potentials in humans: effects of stimulus presentation rate. Electroenceph clin Neurophysiol 1980;49:240–9. Saito T, Yamada T, Hasegawa A, Matsue Y, Emori T, Onishi H, Fuchigami T. Recovery functions of common peroneal, posterior tibial and sural nerve somatosensory evoked potentials. Electroenceph clin Neurophysiol 1992;85:337–44. Shagass C, Shwartz M. Recovery function of somatosensory peripheral nerve and cerebral evoked responses in man. Electroenceph clin Neurophysiol 1964;17:126 –35. Stevens CF, Wang Y. Facilitation and depression at single central synapses. Neuron 1995;14:795 –802. Ugawa Y, Genba-Shimizu K, Kanazawa I. Somatosensory evoked potential recovery (SEP-R) in various neurological disorders. Electroenceph clin Neurophysiol 1996;100:62 –7. Waxman SG. Regional differentiation of the axon: a review with reference to the concept of the multiplex neuron. Brain Res 1972;52: 563– 94. Wegner K, Forss N, Salenius S. Characteristics of the human contra- versus ipsilateral SII cortex. Clin Neurophysiol 2000;111:894–900. Wiederholt WC. Recovery function of short latency components of surface and depth recorded somatosensory evoked potentials in the cat. Electroenceph clin Neurophysiol 1978;45:259–67. Wood CC, Cohen D, Cuffin BN, Yarita M, Allison T. Electrical sources in human somatosensory cortex: identification by combined magnetic and potential recordings. Science 1985;227:1051–3.
Changes in somatosensory evoked responses by repetition of the median nerve stimulation Minoru Hoshiyamaa,b,*, Ryusuke Kakigia a
Department of Integrative Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444-8585, Japan b Department of Health Sciences, Faculty of Medicine, Nagoya University, Nagoya 461-8673, Japan Accepted 30 July 2003
Abstract Objective: We investigate the synaptic factor for the recovery function of evoked responses using a repetitive stimulation technique. Methods: Somatosensory evoked cortical magnetic field (SEF) was recorded following stimulation of the median nerve using single to 6-train stimulation in 8 healthy subjects. The SEF responses after each stimulus in the train stimulation were extracted by subtraction of the waveforms. Results: An attenuation of the SEF components was recognized after the second of the stimuli, but there was no significant attenuation with the third or later stimulations. The root mean square (RMS) of the 1M (peak latency at 20 ms after stimulation) and 4M (70 ms) components were smaller than that of the single stimulation during the train stimulation, while the 2M (30 ms) and 3M (45 ms) components were not attenuated, but the 3M was facilitated at the fourth to sixth stimulation. Conclusion: The synaptic factor was not responsible for the attenuation of the SEF components during repetitive stimulation in healthy subjects. The SEF change disclosed a functional difference among the SEF components during the train stimulation, especially among the later components. q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Somatosensory evoked potential; Somatosensory evoked cortical magnetic field; Repetitive stimulation; Magnetoencephalography; Human; Synaptic depletion
1. Introduction In the study of the recovery function of somatosensory evoked potential (SEP) and cortical magnetic field (SEF), the amplitude of the components decreases as the interstimulus interval (ISI) of the paired stimulation is shortened (Allison, 1962; Meyer-Hardting et al., 1983; Emori et al., 1991; Hoshiyama and Kakigi, 2001, 2002). Previously, we reported that each SEP/SEF component was refractory to a second stimulus after an initial stimulus with a very short ISI (0.5 – 100 ms), and concluded that a structural or functional process of refractoriness existed for high-frequency signals in the primary sensory cortex and there was a difference in refractoriness among the SEP/SEF components (Hoshiyama and Kakigi, 2001, 2002). We speculated, from the results of basic animal studies (Cooley and Dodge, 1966; Fitzhugh, * Corresponding author. Tel.: þ 81-564-55-7779; fax: þ81-564-52-7913. E-mail address: [email protected] (M. Hoshiyama).
1961; Parnas et al., 1976; Waxman, 1972), that one of the factors responsible for the refractoriness to high frequency signals was the morphological structure of the tapering of the axon. However, considering the depletion of the synaptic transmitter following high frequency stimulation, the function of neural conduction including transmission of the signal via synapses might limit the response to the high frequency signals (Stevens and Wang, 1995). Although we considered that the number of synapses was not the major factor for the recovery function of the SEP/SEF components (Hoshiyama and Kakigi, 2001, 2002), there remained the possibility that the difference of synaptic type caused the difference in the recovery function, or the refractoriness to the high frequency signals. We defined the depletion of the synaptic transmitter following high frequency stimulation as ‘synaptic factor’ in the present study. Therefore, the objective of the present study was to investigate the synaptic factor for the recovery function of the responses evoked using a repetitive stimulation
1388-2457/$30.00 q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S1388-2457(03)00285-2
2252
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
technique. The theoretical basis of this study was similar to that of the repetitive stimulation test to investigate the function of the neuromuscular junction, except that we used a multi-channel magnetoencephalography (MEG) system. In addition to the high reliability of dipole estimation, the stimulus artifacts are much smaller in MEG than electroencephalogram recordings, and MEG with axial gradiometers has the advantage of recording cortical responses with a high signal to noise ratio. We used these advantages of MEG for the SEF recording following repetitive stimulation.
2. Methods 2.1. Subjects Eight male volunteers in our department (aged 26 – 39 years) were examined. Informed consent to participate in the study, which was first approved by the Ethical Committee of the National Institute for Physiological Sciences, Okazaki, Japan, was obtained from all participants prior to the study. The subjects were not told about the prospective results of the study prior to the recording. In the present study, the initial 4 components of SEF were investigated (1M, 2M, 3M and 4M, see Section 3). Since the interindividual difference of the later components (3M and 4M) was considerably larger than that of the first two components (1M and 2M), only subjects in whom the 4 components could be identified in a pilot study, participated in the study.
and the left parietal area were mostly covered by the position. 2.3. Stimulus conditions The stimulation was a single to 6-train variety with 3 different ISIs, i.e. 10, 20 and 30 ms. Thus, we used 16 sets of stimuli, which comprised single stimulation, double to 6-train stimulation with a 10 ms ISI, double to 6-train stimulation with a 20 ms ISI and double to 6-train stimulation with a 30 ms ISI. The number of stimuli in the train was expressed with a capital roman numeral followed by the ISI, e.g. V-10 indicated a 5 train stimulation with a 10 ms ISI. Those 16 conditions were controlled by a computer system (Power-1401, CED, Cambridge, UK). The interval between the offset of the last stimulation of a train and the onset of the following train stimulation was 1s (Fig. 1). Each single or train stimulus was randomly delivered to the right median nerve at the wrist using a saddle type electrode. The electrical stimulus was a constant current square-wave pulse of 0.2 ms duration. The intensity of each stimulus was adjusted to produce a slight twitch of the thumb, which was sufficient to produce a SEF of saturated amplitude to a single stimulus. Average intensity was 4.6 mA (range 3.7 –5.3 mA). The duration of the recording epoch was 300 ms after the onset of the train stimulation. The bandpass filter was 1 – 200 Hz with sampling of 2048 Hz.
2.2. SEF recording We focused on the major short- and middle-latency cortical SEF components, i.e. the 4 components appearing approximately 20 ms (1M), 30 ms (2M), 50 ms (3M) and 70 ms (4M) after the stimulation, which were recorded around the central sulcus in the hemisphere contralateral to the stimulated side. The magnetic fields were measured with 37-channel axial-type gradiometers (Magnes, Biomagnetic Technologies Inc., San Diego, CA) in a magnetically shielded room. The detection coils of the system were arranged in a uniformly distributed array in concentric circles over a spherically concave surface. Thus, all of the sensor coils were equally sensitive to the brain’s weak magnetic signals. The device was 144 mm in diameter and its radius was 122 mm. The outer coils were 72.58 apart. Each coil was 20 mm in diameter and the distance between the centers of the coils was 22 mm. Each coil was connected to a superconducting quantum interference device. The measurement matrix was centered at around C3 of the International 10 – 20 system in each subject. The left primary (SI) and secondary (SII) somatosensory cortices
Fig. 1. Experimental procedure and the subtraction method for the SEF analysis. (A) Single to 6-train stimulations with ISIs of 10, 20 and 30 ms were randomly delivered to the median nerve. The number of stimuli in the train was expressed with a capital roman numeral followed by the ISI, e.g. III-30 indicated a 3 train stimulation with an ISI of 30 ms. The interval between the last stimulation and the onset of the first of the following stimulations was 1 s. (B) The SEF waveform evoked by the last stimulation of a train was obtained by subtracting the waveform of a less numerous train stimulation from the waveform, e.g. the SEF waveform evoked by the fifth stimulation (v) was obtained by subtracting the 4-train stimulation (IV) from the 5-train stimulation (V).
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257 Table 1 The peak latency and the RMS values of the SEF components following single stimulation SEF components 1M
2M
3M
2253
with a digital marking in each stimulus train condition. One session took approximately 50 min, and, a short rest was given every 10 min during the recording. The recording was repeated on another day to test the reproducibility of the responses.
4M
Peak latency 18.9 ^ 0.9 28.9 ^ 1.7 45.5 ^ 3.2 72.6 ^ 7.2 (mean ^ SD, ms) RMS 85.5 ^ 30.6 68.1 ^ 38.7 49.0 ^ 19.7 104.4 ^ 32.6 (mean ^ SD, fT)
We setup the recording equipment so as to collect all of the data in one session to minimize the intra-individual variation in the MEG signal. One recording session comprised 200 epochs for each stimulus condition. Two hundred epochs were collected separately for each condition
2.4. Data analysis The MEG signals were collected and stored on a magneto-optical disk, and analyzed for later processing by an off-line system. The epochs of each stimulus condition were collected and averaged separately. The SEF waveform evoked by the last stimulation of a train was obtained by subtracting the waveform of a less numerous train stimulation from the waveform, and the waveform was expressed with a small roman numeral, e.g. the SEF
Fig. 2. The SEF waveform following each train stimulation in one subject. The number of stimuli in the train was expressed as a capital roman numeral, and the arrow indicates the stimulation.
2254
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
waveform evoked by the fifth stimulation in the V-10 was obtained by subtracting IV-10 from V-10, and the waveform was described as v-10. The peak latency of first 4 components of SEF and the root-mean-square (RMS) value at the peak of the components was calculated from the 37-channel waveform. Those values were assessed to analyze each stimulus condition. The peak latency and the RMS values were expressed as a ratio, which was obtained by dividing each value by the value of the SEF following single stimulation (Table 1). The two-way (number of repetition and ISI sequence) repeated measures analysis of variance (ANOVA) was used for statistical analysis of the latency or RMS value. Bonferroni – Dunn’s correction was used for multiple comparisons, when necessary. 3. Results Fig. 2 shows the raw averaged SEF waveforms following each stimulus train recorded from one subject. In the waveforms following the single stimulation, 4 SEF components were identified. The components were termed 1M, 2M, 3M and 4M, and the peak latency and the RMS value at the peak are shown in Table 1. The stimulus artifact was not considerable, which is one of the major advantages of SEF recordings, and subtracted
waveforms were successfully obtained (Fig. 3). The RMS values of the 1M and 4M components attenuated with the repetition of the stimulation in all ISI sequences (P , 0:01), and the attenuation was greater in the 10 ms ISI sequence than in the 20 and 30 ms ISI sequences (P , 0:02) (Fig. 4). However, there was no significant change among the number of repetitions. For the 2M and 3M components, there was evidently a different change in the RMS values. The RMS values of the 2M component did not decline, but tended to increase at iii-10 and ii-20, although the difference was not significant (P ¼ 0:051, ANOVA with Bonferroni – Dunn’s correction) (Fig. 4). For the 3M component, the RMS values increased in the 20 and 30 ms ISI sessions with the number of repetitions (P , 0:02, ANOVA), and the increase at ii-10 was significant (P , 0:02, ANOVA with Bonferroni – Dunn’s correction). The latency was prolonged in general (Fig. 5). The prolongation of the 1M component was significantly greater in the 10 ms ISI sequence than in the 20 and 30 ms ISI sequences (P , 0:02, ANOVA). However, for the 2M, 3M and 4M components, there was no significant difference in the prolongation among the sequences or among the number of stimulus repetitions. The latency ratio of the 3M component was significantly greater than that of 1M in the 20 and 30 ms sequences (P , 0:02, ANOVA).
Fig. 3. The SEF waveform evoked by the last stimulation of a train obtained by the method shown in Fig. 1. The small roman numerals (i –vi) indicate the number of the last stimulus in the train, e.g. the SEF waveform evoked by the last (fourth) stimulation in IV-10 condition is described as iv-10. The subject was the same as in Fig. 2.
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
Fig. 4. The RMS changes of the SEF components during the stimulus repetition. Each value was expressed as a ratio to that of the SEF waveform following single stimulation. The small roman numerals (ii–vi) indicate the stimulus number in the train. The 1M and 4M components were attenuated during the train stimulation, while the 2M and 3M were not. There was no significant decrease in the RMS values in the stimulus train, but for the 3M component, the RMS values rose in the 20 and 30 ms ISI sessions as the number of repetitions increased (P , 0:02, ANOVA), and the RMS values of 3M at the fourth (iv) to sixth (vi) stimulations with ISIs of 20 and 30 ms were greater than those of the second (ii) and third (iii) stimulations (*P , 0:02, ANOVA, Bonferroni–Dunn’s correction).
4. Discussion In the present study, we confirmed that there was an attenuation of the SEF components in terms of amplitude following the second of a train of stimuli but no waning or waxing of the response was recognized between the third and sixth stimuli, and that the pattern of attenuation differed among the SEF components. This is the first report of the change in the evoked response during repetitive stimulation with a very short ISI in humans, although the recovery function of the evoked responses (Allison, 1962; Shagass and Shwartz, 1964; Wiederholt, 1978; Meyer-Hardting et al., 1983; Emori et al., 1991; Hoshiyama and Kakigi, 2001, 2002) and the decrease in SEP during repetitive stimulation with an ISI of more than 500 ms (Angel et al., 1985) were previously reported. Previous studies reported an attenuation of SEP components evoked by stimulation with a relatively high stimulus rate of up to 30 Hz (Pratt et al., 1980; Abbruzzese et al., 1990; Delberghe et al., 1990; Fujii et al., 1994), and that the majority of the loss occurred after the second
2255
stimulation in a repetitive series with an ISI of between 500 and 800 ms (Angel et al., 1985). Saito et al. (1992) and Fujii et al. reported that the mechanism of the attenuation in the recovery function of SEF was partially due to interference between the electrical stimulation and the secondary afferent input from the muscle contraction. Since the muscle contracted during the train stimulation also in the present study and we could not know whether the afferent signal from the muscle was stable or not during the train stimulation, we could not exclude the effects of muscle contraction on the SEF changes, especially on the SEF change following the second stimulation. However, since the decrement of the 1M and 2M components were not affected by the stimulus repetition during the second to the sixth stimulation, we considered that we could focused on another point. We considered that the functional refractoriness of the neuron might partially relate to the synaptic ability to respond to high frequency stimulation, and that the SEP change during repetitive stimulation with a very short ISI would give the information on the synaptic factor of the central nervous system. For example, when the half of the synapses did not respond to the second stimulation, the half of the rest of the synapses and the synapse, which did not respond to the second stimulation, could respond to the third one. The total amount of synapses responded to the third stimulation was 75% of the synapses. Similarly, if the synaptic factor mainly contributed to the attenuation of an SEF component during repetitive stimulation, the response might change in amplitude between the second and later components. The decrement of the 1M component was approximately 50% following the second stimulation at an ISI between 10 and 30 ms, and there was no significant change of the value at the following stimulations. Therefore, the synaptic factor was unlikely to be a major factor in the SEF attenuation during the double and repetitive stimulation, at least at an ISI of 10 –30 ms in healthy subjects, although we could not exclude the partial contribution of the synaptic factor. In the in vitro study using pyramidal neurons in slices of the hypocampal region, a single synapse seemed to require up to 20 ms to refill the transmitter (Stevens and Wang, 1995). The discrepancy between their results and our’s might be explained by that the depression of the transmission for a short ISI was not profound in a group of neurons, as Stevens and Wang (1995) described. However, at an ISI of less than 10 ms, the transmission failure at synapses might occur and cause the attenuation of the SEP/SEF components. In our previous study, we speculated that tapering axons might contribute to the SEF attenuation in the recovery function (Hoshiyama and Kakigi, 2001). We considered a possibility that the morphological structure of axon might responsible for the change of attenuation of 1M in the present study. If the morphological structure was responsible for the SEF attenuation at high frequency stimulation, the attenuation could be stable during repetitive stimulation.
2256
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257
Fig. 5. The latency change during the stimulus repetition. Each value was expressed as a ratio to that of the SEF waveform following single stimulation. The small roman numerals (ii–vi) indicate the stimulus number in the train. The latency was prolonged in general, and the prolongation was significantly greater for the 10 ms than 20 or 30 ms ISI sequence (P , 0:02, ANOVA), and the prolongation of the 3M latency was significantly greater than that of 1M in the 20 and 30 ms sequences (P , 0:02, ANOVA). There was no significant difference in prolongation among the sequences or among the number of stimulus repetitions for 2M, 3M and 4M.
The pattern of attenuation for 1M was evidently different from that for 2M. The 1M components showed a decrease during repetitive stimulation, but the 2M component did not show any significant attenuation. These results were essentially consistent with the results of the previous study (Hoshiyama and Kakigi, 2001). Differences in the generator or structure of the neurons, including axonal structure, were speculated as the reason for the result (Hoshiyama and Kakigi, 2001, 2002), which might cause the functional difference in the response to ‘secondary afferent signals’ (Fujii et al., 1994). With regard to the later responses, the 3M component did not show any significant attenuation during the repetitive stimulation, but it was strengthened at the fourth to sixth stimuli. For the 1M and 2M components, the major generator has been estimated to exist around the primary somatosensory cortex (SI) (Hari et al., 1984; Wood et al., 1985; Kakigi, 1994; Hoshiyama and Kakigi, 2001), but, at the latency of 3M, the activity in the secondary somatosensory cortex (SII) (Karhu and Tesche, 1999; Wegner et al., 2000) and posterior parietal cortex started to overlap (Forss et al., 1994). The SII response was sustained during the repetitive stimulation, while the response in SI was sharp and transient (Forss et al., 2001). The facilitation of 3M by stimulus repetition might be due to the difference in the timing and strength of the inhibition followed by
the excitation (Forss et al., 2001), and, in the present study, sustained muscle contraction during the train stimulation might contributed to the change of the component. On the other hand, the 4M component showed an attenuation in amplitude during the stimulus repetition. The 3M and 4M components could contain the SII activity, but the present results suggested that the function of the 4M component was different from that of the 3M component. Concerning the similarity of the change between the 1M and 4M components during the train stimulation, we could not confirm the mechanism from the present results, although one possibility would be that the 4M response was the secondary response from the 1M signal. The latency of the components was prolonged in general during the repetitive stimulation. The results for the 1M and 2M components were also consistent with the previous results (Hoshiyama and Kakigi, 2001). A prolongation of the latency seemed to be evident for the 3M component. Similarly to the change in amplitude, a functional difference between 3M and the other components was speculated. The jitter of the response might affect more the later components than the early responses, although we are unaware of actual jitter in the SI and SII. Prior to the present study, we tried to record SEP simultaneously with the MEG. However, the electrical artifact from the train stimulation was too large to obtain consistent responses in the subtracted waveform. The correspondence between the SEF and SEP responses should be addressed in a further study using advanced new programs. Abnormal recovery function of SEP has been reported in patients with myotonic dystrophy (Mochizuki et al., 2001), kii amyotrophic lateral sclerosis/parkinsonism-dementia complex (Machii et al., 2003) and subcortical lesions (Ugawa et al., 1996). Those reports disclosed disinhibited cortical excitability in those neurological disorders. The study of the temporal change of SEP/SEF using the method in the present study might give us additional information concerning the cellular and synaptic pathology in those diseases. In conclusion, we reported the change in SEF responses during repetitive stimulation. The attenuation of the amplitude of the SEF component (1M) recognized after the second stimulation did not change with third and later stimulations, but one component (3M) was facilitated by the stimulus repetition. We considered that the synaptic factor was not responsible for the attenuation of the SEF components during repetitive stimulation in healthy subjects. The SEF change disclosed a functional difference among the SEF components during the train stimulation, especially among the later components. References Abbruzzese G, Dall’Agata D, Morena M, Reni L, Trivelli G, Favale E. Selective effects of repetition rate on frontal and parietal somatosensory evoked potentials (SEPs). Electroenceph clin Neurophysiol Suppl 1990; 41:145–8.
M. Hoshiyama, R. Kakigi / Clinical Neurophysiology 114 (2003) 2251–2257 Allison T. Recovery function of somatosensory evoked responses in man. Electroenceph clin Neurophysiol 1962;14:331 –43. Angel RW, Quick WM, Boylls CC, Weinrich M, Rodnitzky RL. Decrement of somatosensory evoked potentials during repetitive stimulation. Electroenceph clin Neurophysiol 1985;60:335 –42. Cooley JW, Dodge FA. Digital computer solutions for excitation and propagation of the nerve impulse. Biophys J 1966;6:583–99. Delberghe X, Mavroudakis N, Zegers de Beyl D, Brunko E. The effect of stimulus frequency on post- and pre-central short-latency somatosensory evoked potentials (SEPs). Electroenceph clin Neurophysiol 1990; 77:86–92. Emori T, Yamada T, Seki Y, Yasuhara A, Ando K, Honda Y, Leis AA, Vachatimanont P. Recovery functions of fast frequency potentials in the initial negative wave of median SEP. Electroenceph clin Neurophysiol 1991;78:116 –23. Fitzhugh R. Impulses and physiological states in theoretical models of nerve membrane. Biophys J 1961;1:445–66. Forss N, Hari R, Salmelin R, Ahonen A, Hamalainen M, Kajola M, Knuutila J, Simola J. Activation of the human posterior parietal cortex by median nerve stimulation. Exp Brain Res 1994;99:309–15. Forss N, Narici L, Hari R. Sustained activation of the human SII cortices by stimulus trains. Neuroimage 2001;13:497–501. Fujii M, Yamada T, Aihara M, Kokubun Y, Noguchi Y, Matsubara M, Yeh MH. The effects of stimulus rates upon median, ulnar and radial nerve somatosensory evoked potentials. Electroenceph clin Neurophysiol 1994;92:518 –26. Hari R, Reinikainen K, Kaukoranta E, Hamalainen M, Ilmoniemi R, Penttinen A, Salminen J, Teszner D. Somatosensory evoked cerebral magnetic fields from SI and SII in man. Electroenceph clin Neurophysiol 1984;57:254– 63. Hoshiyama M, Kakigi R. Two evoked responses with different recovery functions in the primary somatosensory cortex in humans. Clin Neurophysiol 2001;112:1334–42. Hoshiyama M, Kakigi R. New concept for the recovery function of shortlatency somatosensory evoked cortical potentials following median nerve stimulation. Clin Neurophysiol 2002;113:535 –41. Kakigi R. Somatosensory evoked magnetic fields following median nerve stimulation. Neurosci Res 1994;20:165–74. Karhu J, Tesche CD. Simultaneous early processing of sensory input in human primary (SI) and secondary (SII) somatosensory cortices. J Neurophysiol 1999;81:2017 –25.
2257
Machii K, Ugawa Y, Kokubo Y, Sasaki R, Kuzuhara S. Somatosensory evoked potential recovery in kii amyotrophic lateral sclerosis/ parkinsonism-dementia complex (kii AlS/PDC). Clin Neurophysiol 2003;114:564–8. Meyer-Hardting E, Wiederholt WC, Budnick B. Recovery function of short-latency components of the human somatosensory evoked potential. Arch Neurol 1983;40:290–3. Mochizuki H, Hanajima R, Kowa H, Motoyoshi Y, Ashida H, Kamakura K, Motoyoshi K, Ugawa Y. Somatosensory evoked potential recovery in myotonic dystrophy. Clin Neurophysiol 2001;112:793–9. Parnas I, Hochstein S, Parnas H. Theoretical analysis of parameters leading to frequency modulation along an inhomogeneous axon. J Neurophysiol 1976;39:909–23. Pratt H, Politoske D, Starr A. Mechanically and electrically evoked somatosensory potentials in humans: effects of stimulus presentation rate. Electroenceph clin Neurophysiol 1980;49:240–9. Saito T, Yamada T, Hasegawa A, Matsue Y, Emori T, Onishi H, Fuchigami T. Recovery functions of common peroneal, posterior tibial and sural nerve somatosensory evoked potentials. Electroenceph clin Neurophysiol 1992;85:337–44. Shagass C, Shwartz M. Recovery function of somatosensory peripheral nerve and cerebral evoked responses in man. Electroenceph clin Neurophysiol 1964;17:126 –35. Stevens CF, Wang Y. Facilitation and depression at single central synapses. Neuron 1995;14:795 –802. Ugawa Y, Genba-Shimizu K, Kanazawa I. Somatosensory evoked potential recovery (SEP-R) in various neurological disorders. Electroenceph clin Neurophysiol 1996;100:62 –7. Waxman SG. Regional differentiation of the axon: a review with reference to the concept of the multiplex neuron. Brain Res 1972;52: 563– 94. Wegner K, Forss N, Salenius S. Characteristics of the human contra- versus ipsilateral SII cortex. Clin Neurophysiol 2000;111:894–900. Wiederholt WC. Recovery function of short latency components of surface and depth recorded somatosensory evoked potentials in the cat. Electroenceph clin Neurophysiol 1978;45:259–67. Wood CC, Cohen D, Cuffin BN, Yarita M, Allison T. Electrical sources in human somatosensory cortex: identification by combined magnetic and potential recordings. Science 1985;227:1051–3.