Ž .Brain Research 843 1999 161–170www.elsevier.comrlocaterbres

Research report

‘‘Gating’’ of human short-latency somatosensory evoked cortical responsesduring execution of movement. A high resolution electroencephalography

study

Paolo M. Rossini a,b,c, Claudio Babiloni d,), Fabio Babiloni d, Anna Ambrosini d, Paolo Onorati d,Filippo Carducci d, Antonio Urbano d

a IRCCS ‘‘Centro S. GioÕanni di Dio-FBF’’, Ospedale Fatebenefratelli, Ist. Sacro Cuore, Brescia, Italyb A.Fa.R. CRCCS-DiÕisione di Neurologia, Osp. Fatebenefratelli Isola Tiberina, Rome, Italy

c IRCCS ‘‘S. Lucia’’, Via Ardeatina 306, 00179 Rome, Italyd Istituto di Fisiologia umana, 28 Cattedra di Biofisica, UniÕersita’ degli Studi di Roma ‘‘La Sapienza’’, P.le A. Moro 5, 00185 Rome, Italy

Accepted 8 June 1999

Abstract

Ž .The present study aimed at investigating gating of median nerve somatosensory evoked cortical responses SECRs , estimated duringexecuted continuous complex ipsilateral and contralateral sequential finger movements. SECRs were modeled with an advanced highresolution electroencephalography technology that dramatically improved spatial details of the scalp recorded somatosensory evokedpotentials. Integration with magnetic resonance brain images allowed us to localize different SECRs within cortical areas. The workinghypothesis was that the gating effects were time varying and could differently influence SECRs. Maximum statistically significantŽ . Ž .p-0.01 time-varying gating magnitude reduction of the short-latency SECRs modeled in the contralateral primary motor andsomatosensory and supplementary motor areas was computed during the executed ipsilateral movement. The gating effects were stronger

Ž . Ž .on the modeled SECRs peaking 30–45 ms N30–P30, N32, P45–N45 than 20–26 ms P20–N20, P22, N26 post-stimulus. Furthermore,Ž .the modeled SECRs peaking 30 ms post-stimulus N30–P30 were significantly increased in magnitude during the executed contralateral

movement. These results may delineate a distributed cortical sensorimotor system responsible for the gating effects on SECRs. Thissystem would be able to modulate activity of SECR generators, based on the integration of afferent somatosensory inputs from thestimulated nerve with outputs related to the movement execution. q 1999 Elsevier Science B.V. All rights reserved.

Keywords: Gating effect; Somatosensory evoked cortical responses; High resolution EEG; Executed complex sequential finger movements

1. Introduction

Ž .Attenuation gating of median-nerve short-latency so-Ž .matosensory evoked potentials SEPs has been shown

w xduring execution 11,15,30,34,50,52,65 and mental simu-w xlation 13,55 of hand movement ipsilateral to the stimula-

tion. This effect would result from interfering influences ofmotor cortex on cortical, brainstem, and spinal somatosen-

Ž . w xsory relays centrifugal gating 12,55 . Attenuated me-dian-nerve, short-latency SEPs have been also observedduring ipsilateral passive movement or cutaneous sensoryŽ .i.e., hand touch stimulation of the examined hand, due to

) Corresponding author. Fax: q0039-6-4991-0860; E-mail:babilonic@axrma.uniroma1.it

Žinterfering effects of somatosensory inputs centripetal gat-. w xing 10,12,25,34,50 . These data indicate that gating of

SEPs during the executed ipsilateral movement is causedby centrifugal as well as centripetal interference.

Gating effects on different short-latency SEPs duringthe executed ipsilateral movement have not been well-de-

Ž .fined. In traditional electroencephalographic EEG stud-Žies, scalp potentials peaking at about 20 ms parietal N20;

. Ž . ŽNsnegativity , 24 ms frontal N24 , and 32 ms parietal.N32 post-stimulus were gated in some experiments

w x30,35,36 and spared or scarcely gated in other experi-w xments 11,12,55 . Whereas, the potentials peaking at 22 ms

Ž . Ž .P22; Pspositivity and 30 ms frontal N30–parietal P30post-stimulus were found to be markedly gated in all thesereports. During the imagined ipsilateral movement, frontal

0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S0006-8993 99 01716-3

( )P.M. Rossini et al.rBrain Research 843 1999 161–170162

w xN30 was markedly or scarcely attenuated 13,55 . It isplausible that gating of a given SEP component indicates

Ž .interference effects on its cortical generator s .w xThere is a general agreement 1,2,8,9,21,68 that the

Ž .frontal–parietal FP P20–N20 is generated from area 3bof S1. On the contrary, it is still unclear whether the

w xcentral P22 originates from areas 1, 2 of S1 1,8 androrw xfrom M1 20,47 , and whether the frontal cortex partici-

pates with the area 3b to the generation of the frontal N24w xand FP N30–P30 1,2,8,20,49,68 . A frontal contribution

to short-latency SEP generation is supported by reducedŽ .frontal N30 but not parietal P30 in patients with frontalw x w xlobe lesions 28,40–42,51 and Parkinson’s disease 4,51 .

Furthermore, transiently and selectively increased frontalpotentials have been observed in Parkinsonians after a

w xsingle dose of apomorphine 14,53,54 . On the other hand,Ž .in patients with localized lesion of S1 hand area the

frontal–central P20, P22, N30, and N32 were either abol-w x w xished 1 or spared 62,63 . It must be stressed that EEG

potentials are attenuated, spread and overlapped because ofhead volume conduction and electrical reference effectsw x45 . This may explain why previous traditional EEGstudies did not provide conclusive results on gating ofSEPs.

The present study aimed at investigating gating ofmedian-nerve short-latency somatosensory evoked cortical

Ž .responses SECRs , estimated during executed ipsilateralor contralateral complex sequential finger movements. SE-CRs were modeled with an advanced high resolution EEG

Ž .technology see Section 2 that dramatically improvedspatial detail of the original SEPs. This paper reportsunpublished data on the time varying gating effects on theSECRs. From the present experiments, some data address-ing other neurophysiological issues have been reportedpreviously in the form of two short communicationsw x56,66 .

2. Methods

ŽFour normal, right-handed male volunteers age range:.28–37 years participated in the present study. All subjects

gave their written informed consent according to the Dec-laration of Helsinki. The general procedures were ap-proved by the local institutional ethics committee. Theexamined subject was seated in a comfortable recliningarmchair placed in a dimly lit, sound-damped and electri-cally shielded room. Motor tasks concomitant with themedian nerve stimulation consisted of repetitive sequentialopposing movements with the thumb and ulnar fingers ofthe right and left hand. The thumb had to quickly touch theindex finger two times, the middle finger once, the ringfinger three times and the little finger two times and so on

w xin reversed order 13 . Subjects were given two to fourtraining sessions that established approximately stable lev-els of motor performance. They learnt to stare at a visual

fixation point and to minimize blinking and eye move-ments. Surface electromyographic activity of right and left

Žflexor digitorum muscles 1–1000 Hz bandpass; 5000 Hz.sampling rate was recorded to monitor muscle response ofŽ . Žthe operating performance and non-operating involun-

.tary mirror movements hand.ŽSEPs were recorded 1–1000 Hz bandpass; 5000 Hz

.sampling rate by a 128 gold-plated electrode cap. Earlobeipsilateral to the stimulation served as an electrical refer-ence. Electrode impedance was kept lower than 2 kV. All128 electrode 3-D positions were measured with a sonicdigitizer. Nasion, inion and preauricular points were also

w xdigitized for subsequent integration 3 between electrodepositions and subject’s head magnetic resonance imagesŽ .MRIs of the subject’s head. The median nerve wasstimulated at the right wrist with 0.1 ms square wave

Ž .electrical pulses 0.9–1.1 s inter-stimulus intervals . Stimu-lus intensity was adjusted to determine a painless thumbtwitch. The stimulation was performed at rest and duringthe motor execution in pseudo-random order. Electromyo-graphic activity was recorded as in training sessions. Me-dian nerve mixed action potentials were recorded from

Ž .elbow and supraclavear fossa Erb’s point , in order tocheck the steadiness of the signal entering the centralnervous system. Sensory flow towards the central nervoussystem was not influenced by different experimental condi-tions, since the elbow and Erb’s sensory action potentialsremained stable through the whole session. Recording timespanned from 5 ms before to 95 ms after the stimulus. Twoblocks of five hundred artifact-free single trials were aver-

Žaged on-line for each experimental condition waveformsof coupled blocks were quite similar within 50 ms post-

.stimulus . Influence of movement-related cortical poten-w xtials 17 on SEPs recorded during the executed movement

tasks was controlled by analyzing averaged EEG responsesŽ .to the only motor task no median nerve stimulation , with

Žrespect to a 1-Hz trigger delivery no consistent move-.ment-related potentials were observed .

Besides 128 recording channels, the high resolutionw x Ž .EEG technology 3 used in the present study included: i

mathematical correction of head volume conduction andelectrical reference effects by surface Laplacian estimateof the potential over a realistically-shaped, MRI-con-

Žstructed subject’s scalp surface model realistic surface. Ž .Laplacian ; and ii projection of the Laplacian-trans-

formed potentials on the modeled neocortical surface. Scalpand cortical surface models were constructed from 64

ŽT1-weighted sagittal MRIs acquired 1.5 T. Siemens He-.likon System with a 30 ms repetition time, 5 ms echo

time, and 3 mm slice thickness without gap. Before theMRI scan, the nasion, inion, vertex and inter-auricularpoints were labeled using vitamin E pills as markers. TheMRIs were processed with a powerful algorithm that se-lected the points of the contours of scalp, skull, dura mater

w xand cortex 3 . In each sagittal MRI, these contours wereformed by an ordered set of points. The contours extracted

( )P.M. Rossini et al.rBrain Research 843 1999 161–170 163

from contiguous MRIs were triangulated to construct thecompartments simulating scalp, skull, dura mater, andcortical surfaces of the subject’s head. An analytic functionof the scalp surface was obtained by interpolating 577points of this scalp compartment with a 2-D thin plate

w xspline technique 3 . The surface Laplacian was estimatedover the realistic scalp surface model with a 3-D analyticspline. The Laplacian values on some border electrodeswere zeroed since the Laplacian estimation on these elec-trodes may be not reliable. Mean amplitude 256-colourmaps of the Laplacian-derived SEPs were computed on3-D realistically shaped MRIs of the subject’s corticalsurface. The spatial enhancement of the SEPs obtained

Žwith this technology 2–3 cm vs. 6–9 cm spatial resolution.of the traditional EEG allowed an acceptable modeling of

the SECRs.Peak latency of the modeled short-latency SECRs esti-

Žmated within 50 ms post-stimulus FP P20–N20, N24–P24,Ž .N30–P30, frontal–mesial FM N24–N30–P45, and cen-

Ž . .tral–lateral CL P22–N26–N32–P45 cortical responsesw x66 was measured with respect to the time of the stimula-

Ž .tion zerotime . Peak amplitude of the SECRs was refer-enced to the pre-stimulus baseline or the peak of thepreceding cortical response. Since the modeled lateral FPN20–P20, N24–P24, N30–P30, P45–N45 waves se-quence, cortical responses would be generated by tangen-tially oriented dipole sources, plausibly located in the

w xrolandic region contralateral to the stimulation 4,66 . Max-imum amplitude of these responses was obtained by aver-aging maximum Laplacian absolute values computed over

the frontal and parietal cortical regions on which theLaplacian-transformed scalp potentials were projected.Percentage amplitude of short-latency SECRs in variousexperimental conditions were computed in comparison with

Ž .the rest condition rests100% . A log-transformation pro-cedure served to reduce variance and to increase gaussian-

w xity of the data 69 . Maximum differences in log-trans-formed values were selected with a preliminary descriptivedata analysis, while Bonferroni-corrected paired t-test was

Ž .used for the final statistical evaluation p-0.05 .

3. Results

Fig. 1 illustrates mean amplitude 3-D color maps of rawŽ . Ž .left and Laplacian-derived center SEPs of subject 2Ž .S2 , computed over a realistic MRI-constructed subject’sscalp surface model about 20 ms after right median nerve

Ž .stimulation P20–N20 . The scalp Laplacian potential dis-tribution projected on the realistic subject’s cortical surface

Ž .model is also shown right . Compared to the raw scalppotential distribution, the Laplacian-derived cortically pro-jected distribution presented circumscribed frontal P20 andparietal N20 responses closer to the lateral central sulcuscontralateral to the stimulation. This demonstrates that thehigh resolution EEG technology used in the present studymarkedly reduced the head volume conduction effects andprovided a reliable model of the SECRs.

Ž .Fig. 2 plots for another subject S3 the most represen-tative waveforms of short-latency SECRs estimated from

Ž . Ž .Fig. 1. Mean amplitude maps of raw left and realistic Laplacian-derived center EEG activity estimated over a realistic scalp surface model of a subjectŽ . Ž .S2 about 20 ms after right median stimulation at wrist P20–N20 . The Laplacian potential distribution projected on the realistic magnetic resonance

Ž . Ž . Ž .image MRI -constructed subject’s cortical surface model is also presented right . The color scale 256 hues was normalized with reference to theŽ . Ž .maximum value of the potential amplitude. Maximum negativity y100% is coded in blue and maximum positivity q100% in red.

( )P.M. Rossini et al.rBrain Research 843 1999 161–170164

Fig. 2. Most representative waveforms of the short-latency SECRs modeled from the contralateral FL, CL and PL scalp areas and the FM scalp area. TheŽ . Ž . Ž .original potentials were recorded at rest R and during the executed ipsilateral MEI and contralateral MEC motor tasks. N and P indicate negative and

Ž .positive polarity of the potentials, respectively; numbers refer to the response latency with respect to the stimulus zerotime .

Ž . Ž .frontal–lateral FL , FM, CL and parietal–lateral PLŽ .cortically projected scalp surface sites at rest R and

Ž .during the executed ipsilateral MEI and contralateralŽ .MEC movement. No far-field response peaking in the

Ž .range from 14 to 18 ms post-stimulus P14, N18 wasobserved. Response peak at rest had a mean latency ofabout 20, 22, 24, 26, 30, 32, and 45 ms post-stimulus.These peaks were clustered and labeled as FL P20–N24–N30–P45, FM N24–N30–P45, CL P22–N26–N32–P45,and PL N20–P24–P30–N45 based on their cortical surfacedistribution. The FP P20–N20, N24–P24, N30–P30, andŽ .relatively positive–negative P45–N45 cortical responseswere phase-reversed. Compared to response peak latencyat rest, in the other conditions the response peak latencywas substantially unmodified except for a modest reduc-tion in latency for the CL P45 in the executed ipsilateral

Ž . Žtask 44 ms — Rest vs. 41 ms . All SECRs except the CL

.P45 cortical response presented a pronounced attenuationin amplitude concomitant with the executed ipsilateral

Ž .movement Table 1 . In this condition, maximum gatingeffects were registered for the FP N30–P30 cortical re-

Ž 2 .sponse from 4.5 to 0.6 mVrcm . This response wasŽ 2 .slightly enhanced to 4.8 mVrcm , analogously to the FM

Ž 2 .N30–P30 from 1.4 to 1.8 mVrcm , during the executedcontralateral movement. No other remarkable variation ofSECRs was observed.

Fig. 3 depicts 3-D mean amplitude color maps ofshort-latency SECRs modeled from the same subject of

Ž .Fig. 2 S3 at rest and during the executed ipsilateralmovement. In addition, SECR about 30 ms post-stimulusduring the executed ipsilateral movement was shown. Re-garding the rest condition, the P20–N20 map showed adipolar lateral frontal positiverparietal negative responsescontralateral to the stimulation. Compared to the P20–N20

( )P.M. Rossini et al.rBrain Research 843 1999 161–170 165

Table 1Ž .Amplitude of modeled short-latency SECRs estimated from a subject S3

Ž . Ž . Ž .at rest R and during executed ipsilateral MEI and contralateral MECmotor tasks. Amplitude values are expressed in mVrcm2 because SECRswere modeled on the basis of the surface Laplacian estimate, which isproportional to cortical source current density. Labels as in Fig. 2

2Ž .Amplitudes mVrcm

R MEI MEC

P20–N20 2.0 1.3 2.3P22 3.9 1.6 3.1N24–P24 4.4 2.6 3.7N24fm 1.6 0.8 1.0N26 1.5 1.0 1.9N30–P30 4.5 0.6 4.8N30fm 1.4 0.2 1.8N32 4.4 0.9 2.5P45–N45 2.4 0.0 1.7P45fm 1.1 0.4 0.6P45c 2.9 3.6 4.6

map, the N24–P24 and N30–P30 maps showed a dipolarresponse pattern with a similar spatial distribution butopposite polarity, plus a FM negativity with no clear-cutdipolar counterpart. On the other hand, the P22 and P45–N45 maps were mainly characterized by a circumscribedCL unipolar positivity, and the N26 and N32 maps by acircumscribed CL unipolar negativity. The N26 map pre-sented an additional background negativity in the contralat-eral central–parietal area. With respect to the recordedSEPs, the high-resolution EEG technique could modelSECR waves with simultaneous peaks of opposite polarityin frontal and parietal areas and SECR waves that did nothave a ‘‘mirror image’’ of opposite polarity on these areas.Compared to the rest condition, during the executed ipsi-lateral movement the modeled P20–N20, P22, N24–P24and N26 cortical responses had similar topography butreduced amplitude, the modeled N30–P30 and N32 corti-cal responses were completely canceled, and the modeledP45–N45 cortical response was almost unchanged in am-plitude and topography. No consistent change in thesefeatures was observed during the contralateral movement,except that the modeled lateral FP N30–P30 and FM N30cortical responses were slightly enhanced. Similar resultswere obtained from the remaining subjects.

Across-subject SECR peak latencies computed for theexecuted movements were not statistically different from

Fig. 3. Three-dimensional mean amplitude maps of the P20–N20, P22,N24–P24, N26, N30–P30, N32, P45–N45 cortical responses modeled

Ž . Ž .from the same subject of Fig. 2 S3 at rest R and during the executedŽ .ipsilateral movement MEI . In addition, the N30–P30 map for the

Ž .executed ipsilateral movement MEC condition was shown. In the256-color scale, maximum negativity is coded in violet and maximumpositivity in red. Absolute positive and negative values of the percentagecolor scale are those of rest condition.

( )P.M. Rossini et al.rBrain Research 843 1999 161–170166

Ž .Fig. 4. Across-subject mean and S.D. of the amplitude percentage values of the modeled SECRs estimated during the executed ipsilateral motor task MEIŽ .with respect to the rest condition Rs100% . One and two asterisks indicate significance probability levels less than 0.05 and 0.01, respectively.

those measured at rest, which were of 20.1"0.6 for theŽ .modeled P20–N20, 22.2"0.7 ms for the P22, 24.5"0.7ms for the N24–P24, 26.8"1.1 ms for the N26, 30.5"0.9ms for the N30–P30, 33.7"0.8 ms for the N32, and46"1.2 ms for the P45. The only response peak latencyof the modeled P45–N45 was slightly earlier during the

Ž .executed ipsilateral movement 43.8"2.5 ms than at rest.Fig. 4 shows across-subject mean and S.D. of amplitude

percentage values of the SECRs modeled during the ipsi-Ž .lateral motor task rest conditions100% . The mean per-

centage values ranged from about 70% to 85% for themodeled FP P20–N20, N24–P24, FM N24 and CL P22cortical responses. Whereas, these values ranged fromabout 10% to 50% for the modeled FP N30–P30, P45–N45,FM N30, P45, and CL N32 cortical responses. Negligiblereduction in mean percentage values was observed for themodeled CL P45 cortical response. S.D. percentage valuesranged from about 5% to 35% for all modeled SECRs,

Ž .except for the FM P45 cortical response about 78% .Decreased amplitude was statistically more significant forthe modeled FP N30–P30, P45–N45, the FM N30, P45,

Ž .and the CL N32 p-0.01 , than for the FP P20–N20 andŽ .CL P22 cortical responses p-0.05 . Statistically tenden-

tial reduction in amplitude was observed for the modeledFP N24–P24, FM N24, and CL N26 cortical responses.When compared to the rest condition, percentage differ-ences between modeled cortical responses regarded also

Ž .increased FM N30 about 137%; p-0.05 and FL N30–

Ž .P30 about 118%; tendential statistical significance duringthe executed contralateral movement.

4. Discussion

4.1. General remarks

Ž .In the present study, statistically significant p-0.05gating effects on modeled median-nerve short-latency SE-CRs were found during executed complex sequentialthumb–ulnar finger opposition movements. These gatingeffects were not due to a reduced amount of sensory inputsflowing towards the central nervous system, since nosignificant difference in the amplitude of elbow and supra-clavear fossa potentials was found with respect to the restcondition. Coherent gating effects were observed in allfour subjects examined. Thus, statistically significant dif-

Ž .ferences p-0.05 to p-0.01 were achieved with such asmall group of subjects. This is not surprising if somepoints are considered. The intra-individual difference of

ŽSECRs was reduced by using two sets of data mean of.500 trials for each experimental condition. On the other

hand, inter-individual difference of SECRs was minimizedŽwith a markedly increased spatial information content 128

.channels and surface Laplacian estimate and the use ofrealistic MRI-constructed subject’s head model, taking into

( )P.M. Rossini et al.rBrain Research 843 1999 161–170 167

account individual head features. Increased SECR spatialinformation content made unnecessary across subjectsgrand-averaging, which would have overlapped in spaceand time SECR components. The present results wouldindicate that high resolution EEG technologies can modelsomatosensory cortical processes based on a small subjectgroup and could be then useful in clinical environment. Ofnote, in these experimental subjects the same EEG technol-ogy has recently permitted a reliable modeling of corticalresponses to simple voluntary unilateral and simultaneous

w xbilateral one-digit movements 67 . A larger group ofsubjects was not examined because of time consumingprocedures necessary to construct head subject’s scalp andcortex models from MRIs and to model cortical responses

Ž .in the various experimental conditions see Section 2 .The gating effects in the executed movements did not

depend on superimposition of movement-related potentialsw x17 on short-latency SEPs. In fact, the interfering motortask was not associated with gross slow negative shifts, asthose accompanying preparation for a ballistic one-digit

w xmovement 17 . Furthermore, no time correlation betweenŽthe median nerve stimulation used as a reference for the

.averaging procedure and single finger movement canceledmovement-evoked potentials. Of course, the influence ofmovement-related potentials might be more significant on

w xcortical responses later than those examined 64 , and alsoin an experimental condition in which median nerve stimu-lation would have been stimulated during preparation for a

w xvoluntary ballistic one-digit movement 7,35 . It must beremarked that inter-individual variability of cortical re-sponse latency could not be influenced by the spatialattention, since no increase in amplitude of the modeledparietal P30 was seen during the interfering motor taskw x21,26 .

ŽIn the present study, phase-reversed FLrPL P20–N20,. Ž .N24–P24, N30–P30, P45–N45 , FM N24, N30, P45 ,

Ž .and CL P22, N32, P45 SECRs to the median nervestimulation at rest were modeled. Plausibly, the modeled

Ž .lateral FP SECRs were generated tangential dipoles withincentral sulcus cortex. These SECRs anticipated the mod-

Ž .eled CL and FM SECRs generated radial dipoles fromthe crown of the pre- andror post-central gyri and supple-mentary motor area, respectively. An involvement of theM1–S1 in the generation of the FP N30–P30 fits with the

w xresults of previous EEG 9,62 , magnetoencephalographicw x w x8,30,31,68 and subdural recording 1,2 studies and clearlydisagrees with the hypothesis that the frontal N30 isgenerated only or mainly from the supplementary motorarea.

4.2. Cortical gating effects

No modified timing of the modeled SECRs was ob-served during the executed motor tasks as compared to therest condition. A certain anticipation of the CL P45 corti-cal response during the executed ipsilateral motor task was

plausibly provoked by an altered CL N32 cortical re-sponse. Maximum cortical gating effects were associatedwith the executed ipsilateral movement. In this condition,the earliest statistically significant gating involved the FPP20–N20 and CL P22 cortical responses, which wouldsignal the sensory input to the area 3b and areas 1, 2 of S1,

w xrespectively 1,8,20 . The execution of the ipsilateral motorsequence would produce a synchronized repetitive volley

Žfrom muscle spindle involved in pre-synaptic central inhi-.bition and cutaneous afferents, interfering with the affer-

w xent volley originated from the stimulated nerve 11,12 .This would result in a centripetal gating of S1 neurons,which is consistent with the observation that N20 wasgated by somatosensory reafferences due to active orpassive movements but was unaffected by M1 output

w xassociated with motor preparation 11,12,35,36,55 . A cen-tripetal gating of the modeled FP P20–N20 cortical re-sponse is consistent with a gated N20 seen during interfer-

w xing somatosensory stimulation 29,31,32,37,46 and withw xneuromagnetic gating studies 50,55 . No centrifugal gat-

ing of the modeled FP P20–N20 cortical response wouldbe in line with no pre-movement suppressed activity in

w xmonkey area 3b 43 . On the other hand, gating of thiscortical response supports the idea that thumb and ulnarfinger representations in human area 3b are separate but

w xmarkedly interconnected 5,6,33 . Horizontal interconnec-tion among area 3b neurons would favor integrative pro-cessing of separate somatosensory finger afferents at the

w xfirst cortical relay 24,48 . Widely spread horizontal andbidirectional cortico-cortical connections between sensori-motor cortices provide a substrate for muscular coordina-tion that, in turn, allow dynamic adaptation following

Ž . Ž .physiological i.e., learning or pathological i.e., strokew xconditions 18,22,38 .

Ž .During the executed ipsilateral right hand motor task,the gating effects on the modeled FP N30–P30, P45–N45,FM N30 and CL N32 cortical responses were more pro-

Ž .nounced 10–50% with respect to rests100%; p-0.01than those on the modeled FP P20–N20 and CL P22

Ž .cortical responses 70–85%; p-0.05 . Gating of themodeled FP N24–P24 and CL N26 cortical responsesŽ .70–85% was only statistically tendential. No statisticaldifference in gating effects was observed between themodeled FP, FM and CL SECRs, with the exception thatthe modeled FL P45–N45 and FM P45 cortical responses

Ž .were statistically reduced in amplitude p-0.05 and themodeled CL P45 cortical response remained practicallyunchanged. These results extend those obtained in tradi-tional EEG studies in which median nerve stimulation wasdelivered during active and passive finger movements and

w xsustained hand contraction 11,12,34,50,55 . Stable frontalN24 and gated frontal N30 were previously observed with

w xhigh-frequency stimulation rate in healthy subjects 19,27 .ŽFurthermore, the frontal N30 but not the parietal N20 and

.frontal P22, N24 was enhanced due to etomidate adminis-tration and was gated immediately after an electroconvul-

( )P.M. Rossini et al.rBrain Research 843 1999 161–170168

w xsive therapy in psychiatric patients 23 . Also, the frontalŽ .N30 but not frontal P22 was completely abolished during

w xisofluorane anesthesia 44 . These data may suggest thatŽthe later SECRs modeled in the present study N30–P30,

.N32, and P45–N45 were not due to passive repolarizationof the sensorimotor cortical neurons generating the previ-

Ž .ously modeled earlier SECRs P20–N20, P22, N24–P24 .Rather, the later SECRs would signal more advancedstages of an integrative cortical process in which poly-modal somatosensory information from proprioceptive andexteroceptive receptors would be combined with motorprogram information into a polysynaptic cortico-corticalloop responsible for sensorimotor cortical interaction.

Another important finding of the present investigationwas that the modeled FM N30 cortical response increased

Ž .significantly about 137%; p-0.05 and the modeled FPŽN30–P30 cortical response increased tendentially about

. Ž .118% during the executed contralateral left hand motortask. Since SECRs represent post-synaptic activity ratherthan neuronal discharge, these effects could be accountedfor post-synaptic events. It can be speculated that activatedŽ .right M1–S1 and SMA contralateral to the executedmovement produced transcallosal depolarization of neu-

Ž . Ž .rons including pyramidal cells of the opposite leftcortical areas responsible for the generation of the modeledN30–P30 cortical response to the right median nerve

w xstimulation 4,9 . The depolarized pyramidal neurons wouldgenerate recurrent pre-synaptic inhibition that could syn-chronize M1–S1 and SMA neurons receiving directly orindirectly the afferent depolarizing volley. Such an in-hibitory phasing would be based on the presence of in-hibitory recurrent collaterals in cortical inter-neurons,which would inhibit a large number of pyramidal cells. Asa result, the afferent depolarizing volley would influence a

Žlarger number of pyramidal cells generating the modeled.N30–P30 cortical response during the executed contralat-

eral motor task than at rest condition. An alternativeŽ .possibility is that the activation of the right M1–S1 and

SMA contralateral to the motor task would transcallosallyhyperpolarize neurons generating the modeled N30–P30cortical response. This hyperpolarization would decreasetrans-membrane neuronal resistance, thus increasing post-synaptic current flow in response to the afferent depolariz-ing volley. In this case, approximately the same number ofcortical neurons would produce the modeled N30–P30cortical response during the executed contralateral motortask and in the rest condition. Modeled short-latency SE-CRs other than the modeled N30–P30 cortical responsewere unchanged in the contralateral motor task, becausetrancallosal excitatory or inhibitory effects would mainlyregard the polysynaptic and polymodal cortical loop ashaving a special role in the generation of the modeledN30–P30 cortical response. Giant short-latency SEPs havebeen previously recorded during concomitant pathologicalŽ .i.e., cortical reflex myoclonus or physiological activation

w xof the M1–S1 39,57,58,61 . In patients with cortical reflex

myoclonus, paired pulse stimulation determined a periodof 40–200 ms after the first stimulus during which theSEPs following the second stimulus were enhanced in

w xamplitude 16,59 . The afferent volley resulting from thefirst stimulus probably caused an abnormal synchroniza-tion and inhibitory phasing of pyramidal cells generating

w xlater cortical SEPs 59 . Giant cortical SEPs have also beenobserved in normal subjects after an activation of the M1

w xby transcranial magnetic stimulation 39,60 .In conclusion, compared to previous studies on gating

phenomena, the present study permitted to reveal withgreat spatial and temporal details the evolution of theSECRs during the execution of the ipsilateral and con-tralateral finger movements. A maximum statistically sig-

Ž .nificant p-0.05 time-varying gating of the median-nerve short-latency SECRs modeled in the contralateralM1–S1 and the SMA was computed during the executedipsilateral movement. The gating effects were stronger onthe modeled SECRs peaking 30–45 ms than 20–26 mspost-stimulus. The modeled SECRs peaking 30 ms post-stimulus were slightly increased during the executed con-tralateral movement. These results may delineate a dis-tributed cortical sensorimotor system responsible for thegating effects on SECRs. In particular, they provide newevidence in favor of the controversial hypothesis that thefrontal N30 is not due to passive repolarization of thecortical neuronal population generating P20–N20 andN24–P24. Rather, the present study would indicate that thefrontal N30 is related to integrative cortical processescombining polymodal somatosensory and motor programinformation.

Acknowledgements

The authors would like to thank Dr. Luigi Fattorini forcomputer graphics. The research was granted by MURSTof Italy and University of Rome ‘‘La Sapienza’’.

References

w x1 T. Allison, C. McCarthy, S. Wood, J. Jones, Potentials evoked inhuman and monkey cerebral cortex by stimulation of the mediannerve. A review of scalp and intracranial recordings, Brain 114Ž .1991 2465–2503.

w x2 T. Allison, G. McCarthy, M. Luby, A. Puce, D.D. Spencer, Local-ization of functional regions of human mesial cortex by somatosen-sory evoked potential recording and by cortical stimulation, Elec-

Ž .troencephalogr. Clin. Neurophysiol. 100 1996 126–140.w x3 F. Babiloni, C. Babiloni, F. Carducci, L. Fattorini, P. Onorati, A.

Urbano, Spline Laplacian estimate of EEG potentials over a realisticmagnetic resonance-constructed scalp surface model, Electroen-

Ž . Ž .cephalogr. Clin. Neurophysiol. 98 4 1996 363–373.w x4 F. Babiloni, C. Babiloni, F. Carducci, L. Fattorini, C. Anello, P.

Onorati, A. Urbano, High resolution EEG: a new model-dependentspatial deblurring method using a realistically-shaped MR-con-structed subject’s head model, Electroencephalogr. Clin. Neurophys-

Ž .iol. 102 1997 69–80.

( )P.M. Rossini et al.rBrain Research 843 1999 161–170 169

w x5 C. Baumgartner, A. Doppelbauer, L. Deecke, D.S. Brth, J. Zeitl-hofer, W.W. Sutherling, Neuromagnetic investigation of somatotopy

Ž .of human hand somatosensory cortex, Exp. Brain Res. 87 1991641–648.

w x6 C. Baumgartner, A. Doppelbauer, W.W. Sutherling, G. Lindinger,M.F. Levesque, S. Aull et al., Somatotopy of human hand so-matosensory cortex as studied in scalp EEG, Electroencephalogr.

Ž .Clin. Neurophysiol. 88 1993 271–279.w x7 M.B.E. Boecker, R. Forget, C.H.M. Brunia, The modulation of SEPs

during the foreperiod of a forewarned reaction time task, Electroen-Ž .cephalogr. Clin. Neurophysiol. 88 1993 105–117.

w x8 H. Buchner, M. Fuchs, H. Wischmann, O. Dossel, I. Ludwig, A.Knepper, P. Berg, Source analysis of median nerve and fingerstimulated somatosensory evoked potentials: multichannel simulta-neous recording of electric and magnetic fields combined with

Ž . Ž .3D-MR tomography, Brain Topography 6 4 1994 299–310.w x9 H. Buchner, L. Adams, A. Muller, I. Ludwig, A. Knepper, A. Thron,

K. Niemann, M. Scherg, Somatotopy of human somatosensorycortex revealed by dipole source analysis of early somatosensoryevoked potentials and 3D-MR tomography, Electroencephalogr. Clin.

Ž .Neurophysiol. 96 1995 121–134.w x10 D. Burke, S.C. Gandevia, Interfering cutaneous stimulation and

muscle afferent contribution to cortical potentials, Electroen-Ž .cephalogr. Clin. Neurophysiol. 70 1988 118–125.

w x11 G. Cheron, S. Borenstein, Specific gating of the early somatosensoryevoked potentials during active movement, Electroencephalogr. Clin.

Ž .Neurophysiol. 67 1987 537–548.w x12 G. Cheron, S. Borenstein, Gating of the early components of the

frontal and parietal SEPs in different sensory-motor interferenceŽ .modalities, Electroencephalogr. Clin. Neurophysiol. 80 1991 522–

530.w x13 G. Cheron, S. Borenstein, Mental movement simulation affects the

N30 frontal component of the somatosensory evoked potential,Ž .Electroencephalogr. Clin. Neurophysiol. 84 1992 288–292.

w x14 G. Cheron, T. Piette, A. Thiriaux, J. Jacquy, E. Godaux, SEPs at restand during movement in PD: evidence for specific apomorphineeffect on the frontal N30 wave, Electroencephalogr. Clin. Neuro-

Ž .physiol. 92 1994 491–501.w x15 L.G. Cohen, A. Starr, Vibration and muscle contraction affect

Ž .somatosensory evoked potential, Brain 110 1987 451–467.w x16 G.D. Dawson, Investigations on a patient subject to myoclonic

seizures after sensory stimulation, J. Neurol., Neurosurg. PsychiatryŽ .10 1947 141–162.

w x17 L. Deecke, B. Grozinger, H. Kornhuber, Voluntary finger move-ments in man: cerebral potentials and theory, Biol. Cybern. 23Ž .1976 99–119.

w x18 J. DeFelipe, M. Conley, E.G. Jones, Long-range focal collateraliza-tion of axons arising from corticocortical cells in monkey sensory-

Ž .motor cortex, J. Neurosci. 6 1986 3749–3766.w x19 X. Delberghe, D. Mavroudakis, D. Zegers de Beyl, E. Brunko, The

effect of the stimulus frequency on post- and precentral short-latencyŽ .SEPs, Electroencephalogr. Clin. Neurophysiol. 77 1990 86–92.

w x20 J.E. Desmedt, T. Nguyen, M. Bourguet, Bit-mapped color imagingof human evoked potentials with reference to the N20, P22, P27 andN30 somatosensory responses, Electroencephalogr. Clin. Neurophys-

Ž .iol. 68 1987 1–19.w x21 J.E. Desmedt, C. Tomberg, Mapping early SEPs in selective atten-

tion. Critical evolution of cortical conditions used for titrating bydifference the cognitive P30, P40, P100 and N140, Electroen-

Ž .cephalogr. Clin. Neurophysiol. 74 1989 321–346.w x22 J.P. Donoghue, S. Katai, A collateral pathway to the neostriatum

from corticofugal neurons of the rat sensorimotor cortex: an intraŽ .HRP study, J. Comp. Neurol. 201 1981 1–13.

w x23 A. Ebner, G. Deuschl, Frontal and parietal components of enhancedSEPs: a comparison between pathological and pharmacologicallyinduced conditions, Electroencephalogr. Clin. Neurophysiol. 71Ž .1988 170–179.

w x24 E.V. Evarts, J. Tanji, Reflex and intended responses in motor cortexŽ .pyramidal tract neurons of monkey, J. Neurophysiol. 39 1976

1069–1080.w x25 S.C. Gandevia, P. Burke, B. McKean, Convergence in the so-

matosensory pathway between cutaneous afferents from index andŽ .middle finger in man, Exp. Brain Res. 50 1983 415–425.

w x26 L. Garcia-Larrea, H. Bastuji, F. Mauguiere, Mapping study of SEPs`during selective spatial attention, Electroencephalogr. Clin. Neuro-

Ž .physiol. 80 1991 201–214.w x27 L. Garcia-Larrea, H. Bastuji, F. Mauguiere, Unmasking of cortical`

SEP components by changes in stimulus rate: a topography study,Ž .Electroencephalogr. Clin. Neurophysiol. 84 1992 71–83.

w x28 E. Gutling, A. Gonser, M. Regard, W. Glinz, T. Landis, Dissociationof frontal and parietal components of somatosensory evoked poten-tials in severe head injury, Electroencephalogr. Clin. Neurophysiol.

Ž .88 1993 369–376.w x29 C. Hsieh, F. Shima, S. Tobimatsu, S. Sun, M. Kato, The interaction

of the somatosensory evoked potentials to simultaneous finger stim-uli in the human central nervous system. A study using direct

Ž .recordings, Electroencephalogr. Clin. Neurophysiol. 96 1995 135–142.

w x30 J. Huttunen, V. Homberg, Modification of cortical somatosensoryevoked potentials during tactile exploration and simple active andpassive movements, Electroencephalogr. Clin. Neurophysiol. 81Ž .1991 216–223.

w x31 J. Huttunen, S. Ahlfors, R. Hari, Interaction of afferent impulses inthe human primary sensorimotor cortex, Electroencephalogr. Clin.

Ž .Neurophysiol. 82 1992 176–181.w x32 V. Ibanez, M.P. Deiber, N. Sadato, C. Toro, J. Grissom, R.P.

Woods, J.C. Mazziotta, M. Hallett, Effects of stimulus rate onregional cerebral blood flow after median nerve stimulation, Brain

Ž .118 1995 1339–1351.w x33 Y. Iwamura, M. Tanaka, M. Sakamoto, O. Hikosaka, Functional

subdivisions representing different finger regions in area 3 of thefirst somatosensory cortex of the conscious monkey, Exp. Brain Res.

Ž .51 1983 315–326.w x34 S.J. Jones, C. Power, Scalp topography of human SEPs: the effect of

interfering tactile stimuli applied to the hand, Electroencephalogr.Ž .Clin. Neurophysiol. 58 1984 25–36.

w x35 S.J. Jones, J.P. Halonen, F. Shawkat, Centrifugal and centripetalmechanisms involved in the ‘‘gating’’ of cortical SEPs during

Ž .movement, Electroencephalogr. Clin. Neurophysiol. 4 1989 36–45.w x36 S.J. Jones, T. Allison, G. McCharthy, C.C. Wood, Tactile interfer-

ence differences sub-components of N20, P20 and P29 in the humancortical surface somatosensory evoked potential, Electroencephalogr.

Ž .Clin. Neurophysiol. 82 1992 125–132.w x37 R. Kagigi, S.J. Jones, Influence of concurrent tactile stimulation on

somatosensory evoked potentials following posterior tibial stimula-Ž .tion in man, Electroencephalogr. Clin. Neurophysiol. 65 1986

118–126.w x38 J. Kievit, H.G.J.M. Kuypers, Organization of the thalamo-cortical

connections to the frontal lobe in the rhesus monkey, Exp. BrainŽ .Res. 29 1977 299–322.

w x39 T. Kujirai, M. Sato, J.C. Rothwell, L.G. Cohen, The effect oftranscranial magnetic stimulation on median nerve somatosensory

Ž .evoked potentials, Electroencephalogr. Clin. Neurophysiol. 89 1993227–234.

w x40 F. Mauguiere, J.E. Desmedt, J. Courjon, Astereognosis and dissoci-´ated loss of frontal or parietal components of somatosensory evokedpotentials in hemispheric lesions: detailed correlations with clinical

Ž .signs and computerized tomographic scanning, Brain 106 1983271–311.

w x41 F. Mauguiere, J.E. Desmedt, Focal capsular vascular lesions selec-´tively deafferent the prerolandic or the parietal cortex: SEP evi-

Ž .dence, Ann. Neurol. 30 1991 70–175.w x42 F. Mauguiere, V. Ibanez, Loss of parietal and frontal somatosensory`

evoked potentials in hemispheric deafferentation, in: P.M. Rossini,

( )P.M. Rossini et al.rBrain Research 843 1999 161–170170

Ž .F. Mauguiere Eds. , New trends and advanced techniques in clinical`neurophysiology, Electroencephalogr. Clin. Neurophysiol. Suppl. 41Ž .1990 274–291.

w x43 R.J. Nelson, Activity of monkey primary somatosensory corticalŽ .neuron charges prior to active movement, Brain Res. 406 1987

402–407.w x44 M.C. Nogueira, E. Brunko, A. Vandesteen, M. DeRoud, D. Zegers

de Beyl, Different effects of isofluorane on SEPs recorded overŽ .parietal and frontal scalp, Neurology 39 1989 1210–1215.

w x45 P. Nunez, Neocortical Dynamics and Human EEG Rhythms, OxfordUniv. Press, New York, 1995.

w x46 Y. Okajima, N. Chino, E. Saitoh, A. Kimura, Recovery functions ofsomatosensory vertex potentials in man: interaction of evoked re-sponses from right and left fingers, Electroencephalogr. Clin. Neuro-

Ž .physiol. 80 1991 531–535.w x47 N. Peterson, C. Schroeder, J. Arezzo, Neural generators of early

cortical somatosensory evoked potentials in the awake monkey,Ž .Electroencephalogr. Clin. Neurophysiol. 96 1995 248–260.

w x48 B.H. Pubols, W.I. Welker, J.I. Johonson, Somatic sensory represen-tation of fore-limb in dorsal root fibres of raccoon, coati mundi, and

Ž .cat, J. Neurophysiol. 28 1965 312–341.w x49 P.M. Rossini, G.L. Gigli, M.G. Marciani, F. Zarola, M.D. Caramia,

Non-invasive evaluation of input–output characteristics of sensori-motor cerebral areas in healthy humans, Electroencephalogr. Clin.

Ž .Neurophysiol. 68 1987 88–100.w x50 P.M. Rossini, L. Narici, G.L. Romani, M. Peresson, G. Torrioli, R.

Traversa, Simultaneous motor output and sensory input: corticalinterference site resolved in humans via neuromagnetic measure-

Ž .ments, Neurosci. Lett. 96 1989 300–305.w x51 P.M. Rossini, F. Babiloni, G. Bernardi, L. Cecchi, P.B. Johnson, A.

Malentacca et al., Abnormalities of short latency somatosensoryevoked potentials in Parkinsonian patients, Electroencephalogr. Clin.

Ž .Neurophysiol. 74 1989 277–289.w x52 P.M. Rossini, C. Paradiso, F. Zarola, R. Mariorenzi, R. Traversa, G.

Martino et al., Bit-mapped somatosensory evoked potentials andmuscular reflex responses in man: comparative analysis in differentexperimental protocols, Electroencephalogr. Clin. Neurophysiol. 81Ž .1991 454–465.

w x53 P.M. Rossini, R. Traversa, P. Boccasena, G. Martino, F. Passarelli,L. Pacifici et al., Parkinson’s disease and SEPs: apomorphine-in-duced transient potentiation of frontal components, Neurology 43Ž .1993 2495–2500.

w x54 P.M. Rossini, M.A. Bassetti, P. Pasqualetti, Median nerve so-matosensory evoked potentials. Apomorphine-induced transientpotentiation of frontal components in Parkinson’s disease and in

Ž .parkinsonism, Electroencephalogr. Clin. Neurophysiol. 96 1995236–247.

w x55 P.M. Rossini, M.D. Caramia, M.A. Bassetti, P. Pasqualetti, F.

Tecchio, G. Bernardi, Somatosensory evoked potentials during theideation and execution of individual finger movements, Muscle

Ž .Nerve 19 1996 191–202.w x56 P.M. Rossini, F. Babiloni, C. Babiloni, A. Ambrosini, P. Onorati, A.

Urbano, Topography of spatially-enhanced short-latency somatosen-Ž . Ž .sory evoked potentials, NeuroReport 8 4 1997 991–994.

w x57 J.C. Rothwell, J.A. Obeso, C.D. Mardsen, On the significance ofgiant somatosensory evoked potentials in cortical myoclonus, J.

Ž .Neurol., Neurosurg. Psychiatry 47 1984 33–42.w x58 J.C. Rothwell, J.A. Obeso, C.D. Mardsen, Electrophysiology of

Ž .somatosensory reflex myoclonus, Adv. Neurol. 43 1986 385–398.w x59 M. Seyal, Cortical reflex myoclonus. A study of the relationship

between giant SEPs and motor excitability, J. Clin. Neurophysiol. 8Ž .1991 95–101.

w x60 M. Seyal, J.K. Browne, L.K. Hasuoka, A.J. Gabor, Enhancement ofthe amplitude of SEPs following magnetic pulse stimulation of the

Ž .human brain, Electroencephalogr. Clin. Neurophysiol. 88 199320–29.

w x61 H. Shibasaki, Y. Yamashita, Y. Kuroiwa, Electroencephalographicstudies of myoclonus-related cortical spikes and high amplitude

Ž .SEPs, Brain 101 1978 447–460.w x62 J.C. Slimp, L.B. Tamas, W.C. Stolov, A.R. Wyler, Somatosensory

evoked potentials after removal of somatosensory cortex in man,Ž .Electroencephalogr. Clin. Neurophysiol. 65 1986 111–117.

w x63 M. Sonoo, T. Shimpo, K. Takeda, I. Gemba, I. Nakano, T. Mannen,SEPs in two patients with localized lesions of the postcentral gyrus,

Ž .Electroencephalogr. Clin. Neurophysiol. 80 1991 536–546.w x64 M.J. Soso, E.E. Fetz, Responses of identified cells in postcentral

cortex of awake monkeys during comparable active and passive jointŽ .movements, J. Neurophysiol. 43 1980 1090–1109.

w x Ž65 M.C. Tapia, L.G. Cohen, A. Starr, Selectivity of attenuation i.e.,.gating of somatosensory potentials during voluntary movement in

Ž .humans, Electroencephalogr. Clin. Neurophysiol. 68 1987 226–230.

w x66 A. Urbano, F. Babiloni, C. Babiloni, A. Ambrosini, P. Onorati, P.M.Rossini, Human short-latency cortical responses to somatosensory

Ž .stimulation. A high resolution EEG study, NeuroReport 8 15Ž .1997 3239–3243.

w x67 A. Urbano, C. Babiloni, P. Onorati, A. Ambrosini, F. Carducci, L.Fattorini, F. Babiloni, Responses of human primary sensorimotorand supplementary motor areas to internally-triggered unilateral andsimultaneous bilateral one-digit movements, A high resolution EEG

Ž . Ž .study, Eur. J. Neurosci. 40 8 1998 285–289.w x68 C. Wood, L.G. Cohen, N. Cuffin, M. Yarita, T. Allison, Electrical

sources in human somatosensory cortex: identification by combinedŽ .magnetic and potential recordings, Science 227 1985 1051–1053.

w x69 H. Zar, Biostatistical Analysis, Prentice Hall, New York, 1984.