318 Reports the Scalp Topography of the Human Visually Evoked Subcortical Potential. G. F. A

Invest. Ophthalmol. Vis. Sci. 318 Reports March 1980

We are indebted to Dr. G. van Lith, Oogziekenhuis, gested the optic nerve as its origin. Cracco and Rotterdam, and to Dr. D. van Norren, Ooglijdersgast- Cracco3 described early oscillatory potentials at huis, Utrecht, for testing our lens design and for sug- 100 cy/sec recorded from a wide scalp distribution gestions for improvement in the manuscript. of electrodes, referred to earlobe electrodes. Early From the Kliniek voor Oogheelkunde, Rijks-Universi- in 1979 we identified a triphasic positive- teit Groningen, Groningen, The Netherlands. Submit- negative-positive component (msec) in some sub- ted for publication July 2, 1979. Reprint requests: Aart jects at latencies of positive 22 (P22)> negative 27 C. Kooijman, Kliniek voor Oogheelkunde, Rijks-Uni- (N27), and positive 35 (Pas).4 Since it appeared im- versiteit Groningen, Oostersingel 59, 9713 EZ Gronin- portant to delineate this component from both the gen, The Netherlands. scalp-recorded ERG and the VECP, we have car- Key words: ERG, Ganzfeld stimulator, LED light ried out a topographic study of the scalp distribution. Materials and method. Observations were made REFERENCES on 14 normal volunteer subjects (eight male and 1. Thijssen JM, Braakhuis W, Pinckers A, and van Lith six female) ages between 19 and 38 years (mean 26 G: Standardized electro-ophthalmography. In Pro- years). All had visual acuities of 6/6 or better. For ceedings of the 170th meeting of the Netherlands this topographical study, electrodes were placed Ophthalmological Society. Junk, The Hague, 1976, according to the International 10/20 system.5 In p. 185. the first study, an anterior-posterior series of elec- 2. Krakau CET, Nordenfelt L, and Ohmen R: Routine ERG recording with LED light stimulus. Ophthal- trodes at FPZ, F8, T4, T6, and Oz and half-distance mologica 175:199, 1977. electrodes at F8V4 and T4% were used (Fig. 1). For 3. Brodner H: ERG-Test fur die Praxis. Klin Monatsbl the study of the transverse distribution, electrodes Augenheilkd 167:909, 1975. were placed at C4, C6, T4, and T8 and at CP4, CP6, 4. Kooijman AC, Bos LPM, and te Strake, L: Differ- T4&, and T8V!, (Fig. 2). All recordings were made entiation of cone and rod responses in the human with common reference, but the choice of a rela- ERG by means of colour frequency characteristics. In tively inactive reference site was confounded by Proceedings of the Xlth ISCERG Symposium. Doc both the ERG and VECP. Investigation of com- Ophthalmol Proc Ser 4:59, 1974. monly used reference sites shows that the midfrontal (Fz) was affected by the ERG and both the The scalp topography of the human visually earlobe and mastoid were highly active for the P22-N rP35 component. Indeed it is likely that the evoked subcortical potential. G. F. A. 2 active nature of this site6 is probably responsible HARDING AND M. P. RUBINSTEIN. for some so-called early components of the visually Stimulus and analysis parameters have been adjusted to evoked potential. It was found that the vertex site provide optimum conditions for producing and recording (Cz) was relatively inactive at the latency of the the early components of flash visual evoked potentials. A early components and had the added advantage of visual evoked subcortical potential (VESP) of mean la- > > being equidistant from all the electrodes in the tency f 23"^28"^ 34 has been recorded in 86% of subjects. anterior-posterior chain. The triphasic wave was maximal at an electrode position slightly posterior to the RolandicISylvian fissure and to- This site was also used for the transverse topo- pographically separate from the lid electroretinogram graphic study, but since the rules of equal in- and the visual evoked cortical potential. Monocular terelectrode distance are negated, a further refer- stimulation shows bilateral reduction of the amplitude of ence on the anterior neck was used for compari- the VESP, indicating that the wave is independent of the son. The subjects were seated in a dimly lit room, retina and optic nerve and must be arising from a and flash stimulation was delivered by a Grass postchiasmal site. PS22 photostimulator 25 cm from the eyes. The early components, i.e., components before Silver-silver chloride electrodes were affixed with 50 msec latency, of the visually evoked potential in collodion, and the resistance was maintained man are poorly documented because of their below 5 Kohm. A PDP8E computer was used to minute amplitude, intersubject variability, and average the response from each of the eight chan- poor repeatability under the standard experimen- nels of the electroencephalogram recorded on an tal conditions used for eliciting the visual evoked Elema Schonander machine. The analysis time cortical potential (VECP).1 was 100 msec, and the bandpass of the equipment Van Hasselt2 reported an "ear-mastoid" poten- was from 66 to 700 Hz. tial of 10 msec latency which could only be re- To maximize the signal, it was found necessary corded in a small percentage of subjects and sug- to average the response to 500 stimuli delivered at

Downloaded from iovs.arvojournals.org on 10/02/2021 Volume 19 Number 3 Reports 319

- 100 msec FLASH Int. 3939 nits. 500 sweeps

Fig. 1. Topographic distribution of potentials recorded in an anterior-posterior chain of electrodes from frontal to occipital pole, referred to vertex (Cz). The activity of this frontal pole (FPZ) shows the inverted ERG and its oscillatory potentials, and each subsequent site shows the decreasing amplitude of these components. At T4^, however, the VESP consisting of a P24-N30-P34 components appears, and this reduces in amplitude at more posterior sites and almost disappears at the occipital pole, which shows the N39-P48 component of the VECP.

REF 1 REF 2

FLASH INTENSITY 3939nits. 500 SWEEPS 6 FLASHES /SECOND

Fig. 2. Topographic distribution of potentials recorded in transverse chains of electrodes referred to vertex (Ref. 1) and to an anterior neck site (Ref. 2) as a comparator. With either

reference, the amplitude of the VESP is shown to be maximal at (or around) electrode T8V4, i.e., an area level with the cocha but behind the external auditory meatus.

Downloaded from iovs.arvojournals.org on 10/02/2021 320 Reports Invest. Ophthalmol. Vis. Sci. March 1980

BnooJor

Flash Intensity-9661 mte 500 Sweeps 6 Rashes/Second

Fig. 3. During binocular stimulation the VESP at N23-P32 is seen bilaterally at T3!4 and T4,A and equal to approximately 7/x.V peak to peak. When the left eye is occluded, the ERG is reduced from the left scalp electrodes (FP1 and F7Vi), but the VESP is still present at the left (posttemporal, T3i/) as it is on the right (T4v0- Similar findings are obtained when the right eye is occluded with a reduced VESP recorded bilaterally.

high intensity (3939 nits) at a rate of 6/sec. Since the mean amplitude of P22-N27 was 1.09 /xV, and the photostimulator produced a "click" with each that of N27-P35 was 2.05 /xV. discharge, trials included a series with the lamp These waveforms were recorded in 12 out of 14 occluded and a further series with white noise de- subjects. It is clear that this wave was independent livered through earphones. Under the former of both ERG and VECP, and its origin was almost conditions no response of similar latency was certainly subcortical. We have, therefore, termed elicited, but when white noise obscured the click, this triphasic wave the VESP. To clarify whether it the visually evoked subcortical potential (VESP) arises from the optic nerve or a postchiasmal site, was unchanged. we compared the response to binocular and mon- Results. In the anterior-posterior topographic ocular stimulation. On monocular stimulation study, the midfrontal ERG (inverted for compari- (Fig. 3), the ERG on the side ipsilateral to the son) recorded from electrode FPZ was seen to be occluded eye was almost totally abolished. How- widely distributed and recordable at more poste- ever, the VESP was reduced bilaterally to approx- rior electrodes (Fig. 1). Its recorded amplitude imately half the amplitude obtained by binocular gradually diminished until an area of relative inac- stimulation. Such a finding would indicate that the tivity was obtained at electrode T4. A similar phe- VESP is not related to the ERG or optic nerve nomenon occurred with the midoccipital VECP, activity and must be of postchiasmal origin. The which was maximal at Oz and gradually decreased fact that the ipsilateral ERG recorded from the in amplitude when recorded at more anterior skin was not totally abolished is related to the sites. At electrode T

Downloaded from iovs.arvojournals.org on 10/02/2021 Volume 19 Number 3 Reports 321

found that the maximal amplitude of the VESP Mrs. Ann Davies for graphic and secretarial assistance. occurred at a site below the temporal electrode Mr. Neville Drasdo provided frequent advice and en- chain and around a point level with the cocha but couragement. behind the pinna, i.e., around the mastoid pro- From the Department of Ophthalmic Optics, Uni- cess. Since this point was at a maximum interelec- versity of Aston in Birmingham, Gosta Green, Birming- trode distance from the Cz reference, we have also ham, England. Submitted for publication May 23, 1979. compared the findings using an anterior neck ref- Reprint requests: G. F. A. Harding, Dept. of Ophthal- erence. Fig. 2 clearly shows that the amplitude of mic Optics, University of Aston in Birmingham, Gosta the signal is greatest, with either reference, at or Green, Birmingham B47ET, England. around the mastoid. Key words: electroretinogram, reference, subcortical, Discussion. The VESP has probably not been topography, visual evoked potential previously identified for three reasons. 1. Inappropriate stimulus and recording pa- REFERENCES rameters were used.2 From our studies, a large 1. Desmedt JE: Visual Evoked Potentials in Man: New number of responses to high-intensity flash stimu- Developments. Oxford, 1977, The Clarendon Press. lation need to be averaged over a short sampling 2. Van Hasselt P: A short latency visual evoked potential time. recorded from the human mastoid porcess and auri- 2. The confounding nature of the scalp-re- cle. Electroencephalogr Clin Neurophysiol 33:517, corded ERG and its associated oscillatory poten- 1972. tials have not always been recognized. Indeed the 3. Cracco RQ and Cracco JB: Visual evoked potentials in illustrations of some authors appear to show man: early oscillator)' potentials. Electroencephalogr mainly oscillatory potentials derived from an ERG Clin Neurophysiol 45:731, 1978. recorded with very short time constants.3 It is of 4. Harding GFA: A mirror for the brain. Inaugural Lec- interest to note that in our search for a common ture, University of Aston in Birmingham, 1979. 5. Jasper HH: Report on the committee of methods reference for the transverse distribution, both the of clinical examinations in electroencephalography. tip of the nose and the chin were found to be Electroencephalogr Clin Neurophysiol 10:370, 1958. highly active for ERG signals. 6. Lehtonen JB and Koivikko MJ: The use of a non- 3. Some authors have assumed that the mastoid cephalic reference electrode in recording cerebral process is inactive for visually evoked potentials3"7 evoked potentials in man. Electroencephalogr Clin However, other authors agree with us that the Neurophysiol 31:154, 1971. mastoid process is highly active for early compo- 7. Allison T, Matsumiya Y, GoffGD, and GoffWB: The nents.2' 6 Our studies demonstrate this site to be scalp topography of the human visual evoked poten- the point of maximum amplitude of the VESP in tials. Electroencephalogr Clin Neurophysiol 42:185, 1977. an anterior-posterior and transverse direction, ir- respective of reference site. It is certainly not true that the potential is maximal at the vertex.3 The VESP elicited under our standard condi- The McCollough effect in rhesus monkey. tions appears to be a relatively stable phenomenon WILLIAM M. MAGUIRE, GLENN E. MEYER,* present in a larger proportion of subjects (86%) AND JOAN S. BAIZER. than previously reported. Because its lateraliza- After viewing red vertical stripes and green horizontal tion is not changed by the occlusion of an eye, its stripes, the eye subsequently views white vertical stripes origin cannot be prechiasmal, from either the ret- as greenish and white horizontal stripes as pinkish. It has ina or the optic nerve, and must arise from been theorized that this phenomenon, known as the postchiasmal centers. The most obvious choice McCollough effect, is related to long-term adaptatioti of would be the lateral geniculate bodies. The VESP cells tuned for both color and orientation. Such cells may have some clinical utility because at present have been found in the visual cortex of the rhestis mon- the combination of the ERG and VECP recordings key. We asked whether rhesus monkeys, like man, expe- is the best that can be achieved in ophthalmic rience a McCollough effect. Two humans and two rhesus monkeys were adapted by requiring them to fixate a sj)Ot electrodiagnosis and this new technique may allow moving slowly across alternating horizontal and vertical a further degree of differentiation of lesions of the gratings of complementary color. Following adaptation, optic pathway. a test grating whose color changed from red to green or We are most grateful to Miss Kim Lyons for the con- green to red was presented. Humans and monkeys were tinued technical help throughout all stages of this proj- instructed or trained to release a response lever during ect. We are also grateful to Miss Jennifer Pullan and the interval that the grating was white. After adaptation,

Downloaded from iovs.arvojournals.org on 10/02/2021