A new method has been developed for recording short-latency somatosensory-evoked potentials after median nerve stimulation. Negative electrical forces recorded with three orthodiagonal bipolar electrodes in the neck had a direction opposite to that of impulse conduction in the proximal peripheral and cervical somatosensory pathway. Sequential tracings of vectors opposite the electrical forces were made in three-dimensional display, thus reproducing the actual time sequence of electrical events in those structures. Fixed generators such as the subcortical nuclei were also analyzed with this technique, and multiple generators of N13 potential (N13a and N13b) were visualized. This technique may be useful in the functional evaluation of the somatosensory pathway in the cervical cord. Key words: vector somatosensory-evokedpotentials Lissajous trajectory three-dimensional display MUSCLE & NERVE 13:1174-1182 1990

VECTOR SHORT-LATENCY SOMATOSENSORY-EVOKED POTENTIALS AFTER MEDIAN NERVE STIMULATION RYUJl KAJI, MD, PhD and AUSTIN J. SUMNER, MD

AIthough

thc use of‘ soinatoserisoi.y-evoke~l potentials (SE:Ps) has liecoriie ;I clinical roiltilie in neurodiagiiostic laboratories, there are still p o t ) lems to be solved 1ieLi)r-c the techiiique will have the optimum utility in clinical applicxtioii. First, conventional recordings use monopolar montages with “a~tive”;inti “refer-ewe” electrodes, assuming that electrical ac:t.ivities at the i-ekr-crice elet:trode are indifferent t o the generators. Desrnedt arid C;her-ori’ demoiistratecl that the forehead “relei-ewe” electrode is riot ac:tually indifferent to SEP generatoi.s, anti they advocated use 01‘ the iioiicephalic reference electrode placeti in the exti-emitics. Kelerential r-ec:ordingswcre, howcvcr, shown to be influericctl by both the geometry aiid suddcri conductance changes of the volume cond u c t o r . 1 1 ~ 1 4It~ 1is” also prone to interftrence t)y muscle or cardiac artefacts. M o r e importantly, there is n o cviderice that tlie noncephalic refiy-

From the Department of Neurology, University of Pennsylvania (both authors), Department of Neurology, Kyoto University, Kyoto, Japan (Dr Kaji) and Department of Neurology, Louisiana State University Medical Center, New Orleans, Louisiana (Dr Sumner). Address reprint requests to Ryuji Kajl, MD, PhD. Department of Neurology, Kyoto University Hospital, Shogoln, Sakyoku, Kyoto, Japan 606 Accepted for publication January 9. 1990 CCC 0148-639X/90/01201174-09 $04.00 0 1990 John Wiley & Sons, inc

1174

Vector SEP

ence electrode is completely intlitferent t o the gene rat ()r. !’24 Sccontl, potentials rccorcled with conventional monopolar leads may have iiiultiple generators occurring at tlie same tiiiic.“ LJnlike per-ipheral nerves, the central soinatoserisory pathways have parallel organization; the central processes o f pri11iai.y sensory ;txoiis terminate at the seginental ttoisil horii as well ;IS in the dorsal columii nuclei. ‘l’hei-cfore,&ctrical ever1i.s in the dorsal columns or the dorsal coluiiiii niiclei riiay take place a t a time wlicii synaptic potentials a t the segmental dorsal horn still persist. I n fact, this was the case in x i animal study. 14 ‘1’0 solve thc above p-oblems, wc developed a metlioci 01’i)ipoIar recoi-tting 01‘metiian ~ k l ~ s . In ’.~ hipolar montages, both electrocles arc supposed to he active, m d there is no need for an “indiffer-erit” clectr-odc. Bipolar- recording is lcss affkcted by the c.on(lllctanc:e change 01‘ the geomctry of the volume coiidiictoi- than noiicephalic rcfeixmcc rec.ol.t~illg,1.j3lt;.2I .und .. it is suitable IOr picking up conrluctcci voIIeys.”~“~2” Because 01‘ttic smaIIer interelectrode distance, interference d u e to muscle o r cardiac artcfkts is minitrial. Moreover, bipolar SKF recording with three orthodiagoiial axcs in the neck can distinguish I)ctween siiiiultaneously occurring multiple generat.ors (eg, N 13a and N I t3b) by analyzing [he spatial orientation of the genei-ator ciipoIe. I’

MUSCLE & NERVE

D e c e m b e r 1990

In this study, we havc liirther claboratcd the principle of bipolar rewrding and dcvelopd a method of visualihg the progression ol somatosensory conduction through the peripheral nerve, the cervical cord, and the lower brainstcm, sequeurially. The conduction path is represented in a I,issajous' trajectory or the traciri r or instantaneous vect.ors 0 1 electric forces. I y , l k ! h 'I'hc trqjec:tory can he reviewed from any angle in t.lireedimensional space. 'I'he analysis of this vector SEI' is based on the time-honored carcliological technique of the vcctorcardiograni.2 METHODS

Kight normal control su+(:ts (5 woirieii and 3 men; agc 23 to 42 ycars) and 3 patients with circumscribed lesions (1 woman and 2 men) were studied. Thc recording elcctrodes were silver cups 5 mm in dianicter, placed in the upper ccrvical region (u&w vector SEP) and in tlic lower ccrvical region (luww vector SEP) as depicted iii Figure 1.

Square wave electric pulses of 0.1 msec duration werc applicd at 5 Hx. to the median ncrve at h e wrist.. 'I'he cathode was 2.5 cm proximal to the auoclc. Stiiiiulus intensity was adjustcd to produce a minimal thumb twitch. Kight-sided stirnulation was uscd unless specified. Inputs from the 4 chan-

ticls (antero-posterior, lateral, longitudinal, and scalp-neck leads) were amplified with a band pass of 32-1,600 Hz (-3dH), arid were averaged 1,024 to 2,048 times with an analysis time of 20 msec (E1'40, 'I'ECA Corporation, Plcasantville, NY). In each electrode pair, the electrode more posterior, caudal, or proximal to h e stimulation was connected to grid 1 of the amplificr. 'fhe averaged waveforms between 7 to 17 msec after stimulation wcrc transferred from EP40 to Macintosh Plus computer (Apple Computer) for furthcr analysis. The time interval twtween analysis points was 40 pscc aftcr the transler. 'The cortical response from thc scalp was also recorded in clinical cases as described elsewherc.I5 In a computer program (VectcwSEP, Copy-

UPPER

/----'

LOWER

y

,

I;

--

-- --.

x=- ( x

Y=- (Y z=- ( 2

x

FIGURE 1. Principles Of the method of vector SEP. (Left,, Electrode positioning. CVll: the seventh cervical spine, 02: inion, 3: midpoint between CVll and 0 2 , 6: a point in the midline anterior neck at the level of CVII, 9 and 10: points in the lateral neck at the midpoints betweeen CVll and 6. 8: a point in the midline anterior neck at the level of 3, 7 7 and 12: points in the lateral neck at the midpoints between 3 and 8. Planes CVll-9-6-10 and 3-11-8-12 are set perpendicular to the spinal column determined by the surface anatomy, and those electrodes constitute near-ofthodiagonal,x, y, and z axes in upper (above) and lower (below) cervical segments as shown. For more detail refer to ref. 15. (Middle) Digital averaged data from upper or lower axes are transferred to a microcomputer, and reversed in sign. (Right) Reversed-sign data are plotted in 3-D Cartesian axes and displayed on 2-D screen by specifying the viewpoint (Appendix 1).

Vector SEP

MUSCLE & NERVE

December 1990

1175

i-ight b y K. Kaji M U , Phl), 1988, written in BASIC: and compiled), the following operation was performed (Fig. 1). At each time point of' analysis, arriplitiidcs of thc potentials recorded in X (anlero-posterior), Y (lateral), arid 2 (longitudiiial) bipolar leads were plotted in tliree-dimensiorial Cartesian axes, and displayed on the screen by selecting a view point in three-dimensional space.'s 'This point in three-climeiisiorial space indicates the direction and the magnitudc of inst.antaneous ektrical force recorded bipolarly (veclor of electrical force). l'lic electrical force thus reconstructed from volume-conducted tipolarly recorded potentials has a direction opposite to that of' nerve coritiuction as shown bclow (motlel of I i -

polar recording; Fig. 2A-C). 'Therefore, the sign of'the potential was reversed before p l o t t i n g in order to trace the nerve conduction. This process of' plotting and displaying was scquentially repcateti in all the analysis points. Scalp- neck I.cc:ording was simultaneously plotted lor cornparison with the Lissajous trajectory (Fig. 3 ) . A simulation of' bipolar recording was made by a cornp~iter(Macintosh Plus, Apple computer) 11sing BASIC language. 'The right median rierve at the wrist was s&ulaied in O W norinal suhject as dcpicted ahove, anti rnetliari nerve coiii[x)un(l nerve action potentials w e r e recur-dcd bipolarly a t t h e elbow (Fig. 211) t o test the thuory piit hi-ward by the sirnulation.

90

n

-

6oOJd\-.--. . ....:.----.... 30' mC4 T1 segment b y ii metrizaniide ccrvical C I ' . She Iiad analgesic areas over the corresponding dermntonics, but her deep sensibility remained entirely noi-mal. N o fintliiigs suggested ext,cnsion of [tic syrinx above C : 3 or syringohulbia. 3'he scalp- neck atid cortical SEPs showed a normal inter-peak I;ilency hetween N I 3

Case 2 (Figure 48).

Vector SEP

DISCUSSION

'Thc present. study lias shown a n e w approach t o rccord SEPs in ni;in. K y aiialyziiig the spitid orieritations of the electrical forces, it is possit)le t o rcprotluce the actual time sequence of' electrical evcmts frorri the proximal peripheral iicrvc ( N 9 sliif't), Ilirougli the dorsal c:olurrili or the roots (N 1 1 sliift), up to the curvical tlorsal horn or thc dorsal coluiiiii nuclci ( N 1Ya o r N I 3 a + b loop). I n the past, aidysis of the l,issajo\ts Irajet:tory in twoclimensioiial plane has heeti artempicd for cervical'" or cortical" SF.P corripotictits. In this study, w e were able t o correlate the, trajectory with the ini pulse conduction i n t t~ree-ditnensional Spa"".

Although w e have pi-esrrr-ned the gcrierators of SEP components iii the aIialysis of the trajectoi-y, generators of some coriiponents are still in debate. Desmedt anrl Cticron8 sliowed iliat tile N 1:3 potential is a composite of' spinal 1'1 3 and f'l.1 fk-field. l h e spinal P l 3 activity was first demonstrated by using esoptiiigenl eIrctrocIes8 atid h;is a strictly antero-p"stt"riorly oi.ieritcrl generator clipolc in die lower cervical scgiiients. Lat.er, i t was found that t.his potential caii be recox-ded iisiri mteroposterior Iipolar neck electrodcs ( N 1 3 ~ 1 ) . l i15 An

,k,

;tniiiial study suggested t.hat this subco~iipotientis a synaptic pot.enria1 generatcd in the layers of the

MUSCLE & NERVE

December 1990

1179

FIGURE 4. Upper and lower vector SEPs in 3 patients with focal lesions in the peripheral or cervical somatosensory pathway. In the left column, recording electrodes and presumed sites of lesions are illustrated. In the middle, upper and lower vector SEPs are shown. In the right column, scalp-neck recordings (CVll-Fpz) are shown for comparison. At the right upper corner are voltage calibrations as in Figure 3. (A) Case 1, with a distal brachial plexus lesion. (B) Case 2, with cervical syringomyelia. (C) Case 3, with C3-4 compressive myelopathy

cervical ciorsal hoIn w h e r e the priiiiary somatosensory axoiis terminate. 14 1'14 far-field, recorded with iionceplialic I.ef'er-ence elecli.odes, is a potential with ;I latency siinilar to the spinal 1'1 3 ailti has axially or-ientatcd generator dipole.8 Mauguiere and c:olleaguesI7 showed tliat PI 4 lh--fielti was selectively lost in a patient with a transverse lesion n( the lower- 111~'dulln and suggested that its generator was the

1180

Vector SEP

bi-aiiistem rriedial lerririiscal pathway. It was later demonstrated that a t least the carlier subcornporient oL' P14 far-field (PIS) was intact in ;I patient with an cxterisive ]wainstem Iesioii. l 5 1)elestr.eanc1 colleagues 1;. also reported a SEP finding in a patient with an extensive poiitine lesion. 'l'hey interpreted it as an ;ibnornial P I 4 far-field, hut a positive peak with :I la[ency similar t o t h a t of P I 3 was still recor-(led (Fig. 3 of' ref. 6). It appears there-

MUSCLE & NERVE

December 1990

FIGURE 5. Cervical tomography of case 3, showing fusion of C2-C3 vertebrae (Klippel-Feil anomaly) and narrowing (A- P diameter 8 rnm) of the spinal canal at C3-C4.

fore t h a t 1'14 far-field itself h a s niulliplc ger~erators. 'I%e cuneate nucleus seems to he one ol' the generators, since it was shown to have a n open field poterrtid distributioii and ii dipole pardy oriented axially in an animaI mocIrI.'4 KipoIarly recorded N 1111) has ;I latency siniilai- to 1'14 far-lield and ail iixially oriented dipole. Both showed the same beliavior in patielits wil ti foc,al ccr-vical c : o d or the hi.airistem Icsion, aiitl they probably share tlic s;me L)espitc the fact that the Lissajous trajectory ref 1cc ts t 11 e cha 11ging (1 i rection of ne r v e co rid LICt ioi1 , the trajectory is i i o t ideiitical with tlic itctual coiiduction pith. It is a collection 01' points oii which thc instantaneous vector of electrical l'orcc termiiiiltcs. For example, after- the. "1 shift, tlic: trajectory rctiims to near its origin, while the actual conduction p i t t i docs nor rcturn in the rcvelx ctircction (Fig. 3 ) . Keturiiing of the trajectory t o the origin iii~rel y indicates a de(.re;1sr o f the laterally directed clectrical li)rce. I f cadi electrical evciit in the somatosensory coriductioii, such a s ihe dors;rl column volley or the cune;itc potential, takes place discretely without overlapping in riiiic, it is expecred that tlie trajectory wc)ultl return t o the ori-

Vector SEP

gin between thcsc events. However, N I3a+b loop appearcd irnrnediately after- N I 1 shift withoui returning t o the origin. This fintliiig suggests that two successive events m a y considerably overlap in t h c . Since major peaks recordcd i r i SEPs are reflectin< o n l y the fastest c:orductiiig synchronized vo\leyP later components of tlie volley may persist, while the fastest volley initiates synaptic potentials. Clinical studies predicted the site of' concluction dmoriiiality in patients with peripheral xiervc o r die cervical c:ord lesions, but t tie coiiveiitional recordings failed. Extensive noiaceptialic ref'ercnce recordings from tlic scalp and the cervical electrodes'" may IX as sensitive as 0111' riietliott in iliis regard, hut they require more electrodes and m a y be prone to more muscle artifacts than ours due to large iiitcr.-elcctrnde ciistaiices. I n ciise 3, N 1 1 shift generated by tlic dorsal columri volley was lost, hut the latericy of the cortical respoiisc (N20) was witliin normal liriiits. Thesc lindings may seem contradictory. We have demonstrated i n a i i ariirnal study that the v o l u m e conducted potential only reflects the onset 01' s y i i chroiiized discharges at the gcneixtor' site. I 'I Desynclir-oiiized conduction is still possiblc eveii witlioiit genei-ating 1-ecordablc evoked potciitials. 'l'hc coi.tic.al response m a y be initiated b y the fastest cotitluctirig iiiipulscs, eveii when the rest of iinpulses are rransmitted slowly duc to ttesyrichronization. 111 addition tu the increased sensitivity for delcctiiig abnoi-rnality, vector SF,P m a y provide a t o o l lor sei-ial liirictiorial ;issessrrierit of the c.c:rvical c:ord. Its ability t o distinguish potentials from e d i ticrve Lract or nucleus should allow, for example, a cotrip;ii-ison of clorsal column asceiiding volleys belijre and after sirrgery in ;i patient with ccr-vical sporitlylosis. Sincc tlic advent of cornputcr averaging, e v o k e d potentials a r e widely eiriploycd in clinical elcclrotiiagnosis. With thc aid of principlcs already developed foi- nrialyziiig tlic electrocardiogram, wliic:h ;ilso rccords volume-i:onducted 1'0tentials, we have dcmoiistratect refinenicnrs which promise to iniprove tlic diagnostic utility of thesc techniques. APPENDIX 1

Foixiula for t,i-;rIrsforniatioii 0 1 ' ;I 9-dimensional point (X, Y , Z)inlo ;I 2-dimensioiial poiiit (SX, SY) on a display screeii is given bclow (L'rorri ref. 16). 'l'hc eye point of' thc observcr is ;it (p, 0, +) in the spheikal axis, whcre p is the distance from the origin of Cariesian axis, 8 being tlic angle from the

MUSCLE & NERVt

December 1990

1181

x-axis, and

+

being the angle from the z-axis. 1) is the distance of the eye point from the screen.

-

Xe = - X * sin 8 + Y cos 9 Ye = - X . C O S ~ . C O S-+ Y - s i n 0 * c o s + Z * sin Ze = - X . c o s e . s i n + - Y - s i n O - s i n + -

+

%.cos+

SX

-

APPENDIX 2

A potential (E) at a given point P was computed based on a dipolar source hypothesis as follows (from ref. 17).

+

+p

= D XdZe

SY = 1)* Ye&

where K is a constant, m bcing a dipolar moment, 9 being the angle made by the nerve and the line connecting the dipole center and point I?. r is the distance from the dipole to p i n t P.

REFERENCES

I . Allison T, llume Al.: A comparative analysis of shortlatcncy sornatosensory evoked powntials in man, monkey, cat and rat. Exp Neurol 1981;72:592-611. 2. Burch GE, Winsor T: A Primer oj'E&ctrocurdwgruphy, Chs. 1 and 5. Yhiiadelipliia, Lea and Fehiger, 1960. 3. Chiapya KlI: hIvoked PotpntMLr in Clinkul Medkine. New York, Raven, 198.3, p 226. 1. Cracco KQ: Spinal evokcd response: Peripheral nerve stimulation in man. Ekctroenc@ha&gr Clin Neurophpiol 1973;35:379-386. 5. Cracco JB, Chacco RQ, Graziaiii I,]: 'Thc spinal cvoked response in infants and children. N G U T O ~1975;25:31-36. ~Y

6. Delestre F, lnnchampt P, Dubas F Neural generator of Y 1 l far-field somsto.sensory evokcd poteiitial studictl in a patielit with pontine Icsion. h:lectromepldogr Clith NWTOphpriol 1986;65:227-230. 7. Desmedt JE,Chcron G:Central somatoscnsory conductioii in man: neural generators and interpcak lateiwies o f ' tlic far-field componcnis recorded from neck and righi or left scalp arid earlobes. l?'lectrm+Iulo~ Clin Nrurophyswl I980;50:.384-403. 8. Desmedt JK, (;heron ( 2 ~rcvcrtebral(esophageal) recording of subcortical somatoscnwry cvoked potentials in man: h e Spkdl PIS component. and thc dud iiature of tbc spinal generators. ~lettroencephalogrClin [email protected] 1981; 52~257-275. 9. Eisen A, Cracco KQ: Ovcruse of cvoked potentials: (:aution. Neurology 19#3;33:618-621. 10. Emerson KG, Pedley TA: Effect of cervical spinal cord Ic-

sions on early components of the median nerve snmatoscnwry cvaked potential. Narolugy 1986;.56:20-26. 11. Fritb KW, Renstcad 'rJ, D a u k Jk Stationary wdvcs recorded at the shoulder after rncdian iicrve stirnulation. Neurology 1986;36: 1458- 1464. 12. Ino -I', Muoi K: Vector airalysis of auditory brain stem rcsponscs (BSR) in human beings. Arch Olurhirudayagol 1980;2!!6:55-62. 13. Krrji R, hjiwara 'I., Kmicyama M: Short-latency vector SEPs: Dual generator or N 13 potential. Elecfruencephulogr G i n Neuqtdywl 1985;56:S110.

1182

Vector SEP

14. Kaji R,Tanaka

K, Kawaguchi S, McCormick F, Kameyama M: Origin of short-latency somatoscnsnry evokcd potentials to median ncrvc stimulation in ihc cat: comparison of recording montages and effect of laminectomy. Brain 1986;1 OI):443-4titl. 15. Kaji R, Sumner AJ: Hipolar recording of short-latency Mmatoscnwry evoked potentials after rncdian nerve stimulation. Nrurohgy 1987;57:410-418. 16. Kimura J, MiLsudome A, Bcxk DO, Yamada 'r, Dickins ( I S Ficltl distribution ot' antidromidy activated diuiml nerve potcdals: model of far-field rkording. Neurdogy 1985;35:lltj4- 1169. 17. Mauquicrc F, Courjon J, Schott R: Disscriation of early SEP comlmncn~in unilateral tramatic section of the lower medulla. Ann Neurnl 1983;13:309-313. 18. Myers Kk:: Microcnmputm Craphics. Reading, MA, AddisonWesley, l(38'2. 19. Nunez PI.: Elec6ric I&4ls of IIw Hruirr. New York, Oxford Univcrsity Press, 198 I, pp 22-24. 20. Phillips L11, 1)auk JR: Lumhsacral spinal evoked potentials. Nectrohgy 1980;W: 1175- 1183. 21. Stegeman DF. Oosteroni AV, Colon El: Far-field cvoked ptcnt.ial components induccd by a propagating gcncrat~,r:computatiorlal cvidcnce. Ektroen+hulogr CIin Nwro$hp+i,l 1987;67:176- 187. 24. 'l'owlc VI., Munson K, O11it.a 'r, ct al: 'lhree-dimensional human samatosensory cvokeci potentials. Eledromuphalogr Clin Neurophswl 1988;7 I :536-347. 23. Wtmdbury J W Potentials in a volumc conductor, in Ruch T