Electroencephalography and Clinical Neurophysiology, 1977, 4 3 : 7 2 5 - - 7 3 1
725
© Elsevier/North-Holland Scientific Publishers, Ltd.
AVERAGE MULTICHANNEL EEG POTENTIAL FIELDS EVOKED FROM U P P E R AND LOWER HEMI-RETINA: LATENCY D I F F E R E N C E S * D. LEHMAN, H.P. MELES and Z. MIR **
Neurologische Universit~'tsklinik, 8091 Zurich (Switzerland)
(Accepted for publication: April 20, 1977)
Visually evoked average potentials recorded from a given combination of scalp electrodes show differing wave forms as a function of retinal stimulus location. Particularly striking are the differences c o m m o n l y obtained from occipital electrodes referred to precentral or ear electrodes, when the upper or lower hemiretina is stimulated with a checkerboard pattern: the resulting wave forms in the t w o conditions are often approximate mirror images (e.g. Les~vre 1973). Halliday and Michael (1970), Michael and Halliday (1971) and Jeffreys and Axford (1972) suggested anatomical reasons for this finding, i.e. that orientation and location of the active elements in the skull may best be modelled b y an equivalent dipole of near-inverted polarity in the t w o conditions. Recently, the latency of visually evoked averaged EEG potentials has b e c o m e a useful tool in the diagnosis of multiple sclerosis (Halliday et al. 1972, 1973), employing centre field stimulation with a reversing checkerboard pattern as the usual procedure. In the majority of cases, stimulation of the lower hemiretina evokes scalp potentials of smaller amplitude than stimulation of the upper * Supported by Swiss National Science Foundation, EMDO--Stiftung, Ziirich, Hartmann-Miiller-Stiftung, Ziirich and Smith--Kettlewell Institute, San Francisco, Calif.,U.S.A. ** Fellow fo the Roche-Studienstiftung, Basel. Present address: Draskoviceva 4A, 4100 Zagreb, Yugoslavia.
hem[retina (Les~vre 1973; Spekreijse et al. 1973; Cappin and Nissim 1975; Lehmann and Mir 1976), presumably because of the more distant location of the generator area (e.g. Michael and Halliday 1971). Therefore, potentials evoked during fixation of the stimulus field centre (summing upper and lower hemiretinae responses) will largely reflect properties of the upper hem[retinal response. In the latter case, the latency of the largest occipital positive wave during the first 250 msec is slightly shorter in comparison with responses to centre field fixation (Lehmann and Mir 976). In a few subjects, lower hemiretinae stimulation evokes large responses, causing late peak times of the occipital positivity during centre stimulus field fixation, thus giving rise to false positive clinical results which may be avoided b y upper hem[retinal stimulation (Lehmann and Mir 1976). In this context it appears important to clarify the cause of the approximate inversion of polarity which is consistently observed between the potential wave forms evoked b y upper and b y lower hem[retinal stimulation. In the present paper we investigate these potential fields and conclude that the two stimuli evoke fields which differ distinctly in latency and somewhat in topography, b u t are not inverted in polarity (for preliminary report see Lehmann et al. 1976). The data used in this paper agree with comparable results of other workers (e.g., Michael and Halliday 1971).
726
D. LEHMANN ET AL.
Methods A healthy volunteer (28 years of age) was seated c o m f o r t a b l y in a sound-attenuated dark r o o m , with his head in a chin-forehead rest. The checkerboard pattern (50 min arc checks, contrast 96%, illumination 200 ft L. for white and 8 ft L. for dark checks) was backprojected o n t o a translucent screen covering a circular target of 16 deg arc in the centre of a 60 deg black panel. The checkerboard pattern was reversed within 6 msec at 2/sec by a mirror m o u n t e d on a galvanometer. Triangular 30 min fixation marks were attached at 30 min b e y o n d the upper and lower edges o f the target area. Because of the circular stimulus, small segments o f the hemiretina near the horizontal meridian were unstimulated. We chose this arrangement in order to reduce accidental stimulation o f the wrong halffield to a minimum in case o f slippage of fixation. Actually, differences between responses to rectangular and to circular hemiretinal stimuli were very minor.
I
R B
R-A
UPPER
HEMI-RETINA
At 4 min intervals, the subject was asked to put his head into the headrest, close one eye with a shutter attached to the headrest, and to fixate the upper or lower mark. 10 sec after he report ed 'ready' via the intercom, data collection in 39 simultaneous channels was started, and ended after 140 presentations (one run). The subject's face and eye were observed via closed circuit TV during data collection. Repeated runs yielded very similar wave forms. The example examined in this paper is the third run for upper and lower fixation with the right eye. Thirty-nine Grass gold plated cup electrodes were applied to the scalp with Grass paste, at about equal distances (4 cm) over an area between inion and about 70% o f the inion--nasion distance, and symmetrically covering 70% of the distance between the meati acustici externi. The electrode array is indicated in the inset of Fig. 2. Data from 2 anterior temporal electrodes were om i t t ed because of muscle artefacts. Using a 48-channel data handling system (constructed by J.M. Madey
R C
t!
/
j
J ~%v
/
i
0
LOWER
!
HEMI RETINA I
103
I
146 MSEC
Fig. 1. Average potential wave forms (n = 140) evoked by checkerboard reversal shown to the upper (above) and lower (below) hemiretina. Records between 3 active midline electrodes ('C' at inion, 'B' at 12%, and 'A' at 24% of inion--nasion distance) and midline reference ('R', at about 30% nasion--inion distance). Positivity of active electrodes represented by downward deflection. Zero potential difference indicated by dotted line. Times of maximal amplitudes of occipital positivity in the 2 stimulus conditions marked by vertical lines. Analysis time 256 msec, amplitude calibration 6.5 pV. From same data as Figs. 2 and 3.
HEMIRETINA-EVOKED POTENTIALS
727
and V. Corti) the EEG data were pre-amplifled (50% down at 0.5 and 100 c/sec, 12 dB per octave), sampled at 750 s/sec/channel, and multiplexed in groups of 8 channels for AM recording on 6 tracks of an instrumentation tape recorder. Off-line demultiplexing, A-D conversion and averaging in a laboratory computer were followed by calibration and transformation of the data into sequences of maps of potential distributions on the scalp in a large computer system (Lehmann 1971, 1975). Evoked potential wave forms, as conventionally recorded between electrode pairs, were reconstructed from the multichannel data stored in the large computer system (samples in Fig. 1).
Results Fig. 1 illustrates 3 conventional evoked potential wave forms reconstructed from the multichannel data, for the 2 stimulus conditions: the potentials show the weU-known approximate inversion of polarity between responses to upper and lower hemiretinal stimulation recorded between occipital and anterior electrodes. From the wave forms evoked by upper hemiretinal stimulation, the latency of the largest positive component was determined (103 msec, record R-B, upper part of Fig. 1), which more or less coincided in time with the first prominent negative peak evoked by lower hemiretinal stimulation.
UPPER HEMI-RETINA
LOWER HEMI-RETINA
88
103
118
132
146
161 MSEC
Fig. 2. Average (n = 140) scalp field distributions evoked by upper (above) and lower (below) hemiretinal stimulation with checkerboard reversal, at different times (indicated in msec) after stimulus reversal. The electrode array is schematically illustrated in the inset. The field values are computer-interpolated between electrode sites, and mapped as equipotential line plots. Each line indicates a 1 pV step. Positive field maxima indicated by plus signs, negative maxima by minus signs. Note that the location, and thus the potential value, at the c o m m o n reference electrode is o f no consequence for individual maps. The right anterior positivity at 103/118 msec for upper, and 146/161 msec for lower hemiretinal stimulation is not contingent upon the eye stimulated: it was observed with right eye and with left eye stimulation. F r o m same data as Figs. 1 and 3.
728
D. LEHMANN ET AL.
Similarly, the latency of the largest positive c o m p o n e n t evoked by lower hemiretinal stimulation (146 msec, record R-C, lower part of Fig. 1) about corresponded in time with a negative peak evoked by upper hemiretinal stimulation. Fig. 2 illustrates the scalp field distributions evoked by the 2 stimuli at the 2 peak latencies, and at additional times at 15 msec intervals immediately before, between and after the peak latencies. At 88 msec after the stimulus there is an occipital positivity in both stimulus conditions, clearly visible somewhat anterior to the inion during upper stimulation, b u t also detectable at the inion during lower stimulation. At 103 msec, the occipital positivity evoked by upper stimulation reaches its peak and becomes less prominent in subsequent maps. The map sequence evoked b y lower stimulation shows continuing build-up of the occipital positivity, which reaches its maximum at 132 and 146 msec. The maps obtained at 103 msec for upper and at 132 and 146 msec for lower stimula-
tion are strikingly similar. Thus, the fields evoked b y lower stimulation showed occipital peak positivities 43 msec later than those evoked by upper stimulation. A chronotopogram (R~mond 1968) is an efficient data display for the survey of crucial features of field distribution in map sequences. If we consider only the scalp potential values obtained in our 2 stimulus conditions from the 7 midline electrodes at 15 msec intervals, the chronotopograms shown in Fig. 3 result. Obviously, the shapes of the chronotopograms produced by upper and lower hemiretinal stimulation are very similar - - except that they are shifted in time by 43 msec. The conclusion is again that not only during, but also preceding and following the times of maximal amplitude differences in the fields {maximal 'hillyness'; Lehmann 1971), the field distributions evoked b y the 2 conditions are quite similar (see also Fig. 2), strengthening the interpretation of a delayed response evoked by lower hemiretinal stimulation.
UPPER HEMI RETINA 43
58
73
88
103
118
132
146
161 MSEC
132
146
161
175
191
205 MSEC
ANTERIOR
INION LOWER HEMI RETINA 88
103
118
ANTERIOR
INION
Fig. 3. Chronotopograms of average EEG potentials evoked by checkerboard reversal shown to upper and lower hemiretina. Illustrated are potential values along a midline row of 7 electrodes (vertical) at indicated times after stimulus reversal, in 15 msee intervals (horizontal). Pcaitivity white, negativity hatched; dotted line connects maximal positive values; solid line connects maximal negative values. Isopotential lines at 0/~V, which is mean value of the voltages recorded from all 37 electrodes at the given moment (average reference). Note similarity of c h r o n o t o p o g r a m s , b u t also shift in t i m e after stimulus, b e t w e e n t h e t w o conditions. F r o m same data as Figs. 1 and 2.
HEMIRETINA-EVOKED POTENTIALS Discussion Obviously, the approximate wave form inversion o f responses to upper and lower hemiretina stimulation in conventional records comes a b o u t by a time shift of the positive response m a x i m u m to lower stimulation which makes it coincide more or less with the subsequent negative response maxim u m to upper stimulation. Minor optical differences of light beams entering the eye at different angles cannot explain the wave form differences in the 2 conditions, since wave forms remain constant over a wide range of stimulus parameters (contrast sharpness, light intensity: Lehmann and Mir 1976). The present paper offers new interpretations based on topographical evaluations which became possible because of more extensive data collection. Our data -- as far as electrode placements are comparable - - a r e in agreement with those of other studies using similar stimuli (e.g. Michael and Halliday 1971). The data collected below the inion b y these authors confirm the positive field peaks at about the inion which our Fig. 2 indicates at 88 and 103 msec for lower hemiretinal stimulation. Our conclusion that the different wave forms reflect responses of different latencies, and not of inverted polarities, is in agreement with a report b y Eason et al. (1967) w h o found longer latencies of EEG responses evoked by xenon flashes to the lower hemiretina than to the upper hemiretina. In fact, the observed wave form inversion in the two conditions is often only very approximate, i.e. sometimes the latency differences are small (Lehmann and Mir 1976). In any case, differences between conventionally recorded EEG waves (simultaneous recordings, or different conditions) require careful evaluation, since they may be caused by the field values at the reference electrode (Lehmann 1975). An example of the influence o f the potential value at the reference electrode is the negative peak at 103 msec in our conventionally recorded wave forms evoked b y lower hemi-
729 retinal stimulation: at this time, the field value at the reference 'R' of Fig. i was somewhat more positive than the values at 'A' and 'B' (Figs. 1 and 2) at the ascending slope o f the occipital positivity. The wave form R-C of Fig. 1 (lower row) illustrates that a negative peak of a conventional wave form may actually be a near-zero difference value which is preceded and followed b y positive differences. Besides these electrophysiological data, there are classical anatomical, perceptual and behavioural findings which indicate that the human upper hemiretinal system is more powerful than its lower counterpart (see pp. 327--328 in W o o d w o r t h 1938). The upper human hemiretina has more rods and cones per mm 2 than the lower half (Osterberg 1935). There are fewer intracortical projections from lower than from upper hemiretinal cortical areas (Zeki, personal communication). Visual acuity is better for the upper hemiretina (Landolt and Hummelsheim 1904), in agreement with the assumption that denser receptor packing correlates with higher acuity (ten Doesschate 1946). In addition, m o t o r reaction time to light flashes in the upper hemiretina is shorter than in the lower (first reported by Hall and von Kries 1879). In which way the different delays of the evoked responses relate to these data will have to be clarified. At any rate, it could be argued that phylogenetically, human enemies and prey were expected in the lower half of the visual field and therefore, improved performance of the upper hemiretinal system was advantageous.
Summary Checkerboard reversals shown to the upper hemiretina evoke EEG potentials which in anterior-posterior derivations are approximately inverted in polarity compared with potentials evoked by lower hemiretinal stimulation. Using a 48-channel system, EEG scalp
730
field distributions of such responses were mapped. Examination of the map sequences shows that occipital positive maximal field values start to develop at about the same time in the two stimulus conditions, b u t peak much earlier and somewhat more anteriorly for upper than for lower hemiretinal stimulation. Thus, there is a difference of response latency in the two conditions, which accounts for the approximate inversion of polarity. Possible correlations with reports of higher receptor cell density, higher visual acuity, and shorter m o t o r reaction time of the upper hemiretinal system are noted.
Rdsumd Repdrage par canaux multiples des champs de potentiel EEG moyens obtenus d partir des demi-rgtines supgrieure et infdrieure. Diffdfences de latences Une stimulation par damier de la demir~tine sup~rieure ~voque des rdponses EEG qui, en d~rivation antdro-post~rieure, sont pratiquement de polarit~ inverse de celles ~voqu~es par stimulation de la demi-rdtine inf~rieure. La distribution des champs sur le scalp correspondent ~ ces rdponses a dtd dtablie l'aide d'un syst~me ~ 48 canaux. L'examen des cartes montre que les maxima de positivit4 commencent au m~me m o m e n t dans les deux conditions de stimulation, mais que le sommet est atteint plus t6t, et est plus ant~rieur, pour la stimulation sup~rieure que pour celle de la demi-r~tine inf~rieure. Une difference de latence de r~ponse apparaft ainsi entre les deux conditions, ce qui rend compte de l'inversion approximative des polarit~s. On associe ces r~sultats ~ ceux qui signalent une plus grande densitd de r~cepteurs, une plus grande acuit~ visuelle, dans la demi-rdtine sup~rieure, ainsi qu'un temps de r~action visuomoteur inf~rieur ~ partir de sa stimulation.
D. LEHMANN ET AL.
References Eason, R.G., White, C.T. and Oden, D. Averaged occipital responses to stimulation of sites in the upper and lower halves of the retina. Perception Psychophysics, 1967, 2 (10): 423--425. Cappin, J.M. and Nissim, S. Pattern visual evoked responses in the detection of field defects in glaucoma. Arch. Ophthal., 1975, 93: 9--18. Jeffreys, D.A. and Axford, J.G. Source locations of pattern-specific components of human visual evoked potentials. II. Component of extrastriate cortical origin. Exp. Brain Res., 1972, 16: 22--40. Hall, G.S.. und yon Kries, J. Ueber die Abh~/ngigkeit der Reaktionszeit vom Ort des Reizes. Arch. Anat. Physiol. (Leipzig), 1879, Suppl.: 1--10. Halliday, A.M. and Michael, W.F. Changes in patternevoked potentials in man associated with the vertical and horizontal meridians of the visual field. J. Physiol. (Lond.), 1970, 208: 499--513. Halliday, A.M., McDonald, W.I. and Mushin, J. Delayed visual evoked response in optic neuritis. Lancet, 1972, 1 : 982--985. Halliday, A.M., McDonald, W.I. and Mushin, J. Visual evoked response in diagnosis of multiple sclerosis. Brit. Med. J., 1973, 4: 661--664. Londolt, E. und Hummeisheim, E. Die Untersuchung der Funktionen des excentrischen Netzhautgebietes. In T. Saemisch (Ed.), Graefe-Saemisch Handbuch der Gesamten Augenheilkunde, Vol. 4. Engelmann, Leipzig, 1904: 503--583. Lehmann, D. Multichannel topography of human alpha EEG fields. Electroenceph. clin. Neurophysiol., 1971, 31: 439--449. Lehmann, D. EEG phase differences and their physiological significance in scalp field studies. In E. Dolce and H. Kiinkel (Eds.), Computerized EEG analysis. Fischer, Stuttgart, 1975 : 102--110. Lehmann, D. und Mir, Z. Methodik and Auswertung visuell evozierter EEG-Potentiale bei Verdacht auf multiple Sklerose. J. Neurol., 1976, 213: 97--103. Lehmann, D., Meles, H.P. and Mir, Z. Scalp field maps of averaged EEG potentials evoked by checkerboard inversion. Biomed. Technik, 1976, 21 (Suppl.) 117--118. Les~vre, N. Potentiels ~voqu~s par des patterns chez l'homme: influence de variables caracterisant le stimulus et sa position dans le champ visuel. In A. Fessard et G. Lelord (Eds.), Activit~s ~voqu~es et leur conditionnement. INSERM, Paris, 1973: 1--22. Michael, W.F. and Halliday, A.M. Differences between the occipital distribution of upper and lower field pattern evoked responses in man. Brain Res., 1971, 32: 311--324. ¢}sterberg, G. Topography of the layer of rods and
HEMIRETINA-EVOKED POTENTIALS cones in the human retina. Acta ophthal. (Kbh.), 1935, Suppl. 6: 1--102. Rdmond, A. The importance of topographic data in EEG phenomena, and an electrical model to reproduce them. Electroenceph. clin. Neurophysi'ol., 1958, Suppl. 27: 27--49. Spekreijse, H., van der Tweel, L.H. and Zuidema, T.
731
Contrast evoked responses in man. Vision Res., 1973, 13: 1577--1601. Ten Doesscha~, J. Visual acuity and distribution of percipient elements on the retina. Ophthalmologica (Basel), 1946, 112: 1--18. Woodworth, R.S. Experimental psychology. Holt, N e w York 1938.
725
© Elsevier/North-Holland Scientific Publishers, Ltd.
AVERAGE MULTICHANNEL EEG POTENTIAL FIELDS EVOKED FROM U P P E R AND LOWER HEMI-RETINA: LATENCY D I F F E R E N C E S * D. LEHMAN, H.P. MELES and Z. MIR **
Neurologische Universit~'tsklinik, 8091 Zurich (Switzerland)
(Accepted for publication: April 20, 1977)
Visually evoked average potentials recorded from a given combination of scalp electrodes show differing wave forms as a function of retinal stimulus location. Particularly striking are the differences c o m m o n l y obtained from occipital electrodes referred to precentral or ear electrodes, when the upper or lower hemiretina is stimulated with a checkerboard pattern: the resulting wave forms in the t w o conditions are often approximate mirror images (e.g. Les~vre 1973). Halliday and Michael (1970), Michael and Halliday (1971) and Jeffreys and Axford (1972) suggested anatomical reasons for this finding, i.e. that orientation and location of the active elements in the skull may best be modelled b y an equivalent dipole of near-inverted polarity in the t w o conditions. Recently, the latency of visually evoked averaged EEG potentials has b e c o m e a useful tool in the diagnosis of multiple sclerosis (Halliday et al. 1972, 1973), employing centre field stimulation with a reversing checkerboard pattern as the usual procedure. In the majority of cases, stimulation of the lower hemiretina evokes scalp potentials of smaller amplitude than stimulation of the upper * Supported by Swiss National Science Foundation, EMDO--Stiftung, Ziirich, Hartmann-Miiller-Stiftung, Ziirich and Smith--Kettlewell Institute, San Francisco, Calif.,U.S.A. ** Fellow fo the Roche-Studienstiftung, Basel. Present address: Draskoviceva 4A, 4100 Zagreb, Yugoslavia.
hem[retina (Les~vre 1973; Spekreijse et al. 1973; Cappin and Nissim 1975; Lehmann and Mir 1976), presumably because of the more distant location of the generator area (e.g. Michael and Halliday 1971). Therefore, potentials evoked during fixation of the stimulus field centre (summing upper and lower hemiretinae responses) will largely reflect properties of the upper hem[retinal response. In the latter case, the latency of the largest occipital positive wave during the first 250 msec is slightly shorter in comparison with responses to centre field fixation (Lehmann and Mir 976). In a few subjects, lower hemiretinae stimulation evokes large responses, causing late peak times of the occipital positivity during centre stimulus field fixation, thus giving rise to false positive clinical results which may be avoided b y upper hem[retinal stimulation (Lehmann and Mir 1976). In this context it appears important to clarify the cause of the approximate inversion of polarity which is consistently observed between the potential wave forms evoked b y upper and b y lower hem[retinal stimulation. In the present paper we investigate these potential fields and conclude that the two stimuli evoke fields which differ distinctly in latency and somewhat in topography, b u t are not inverted in polarity (for preliminary report see Lehmann et al. 1976). The data used in this paper agree with comparable results of other workers (e.g., Michael and Halliday 1971).
726
D. LEHMANN ET AL.
Methods A healthy volunteer (28 years of age) was seated c o m f o r t a b l y in a sound-attenuated dark r o o m , with his head in a chin-forehead rest. The checkerboard pattern (50 min arc checks, contrast 96%, illumination 200 ft L. for white and 8 ft L. for dark checks) was backprojected o n t o a translucent screen covering a circular target of 16 deg arc in the centre of a 60 deg black panel. The checkerboard pattern was reversed within 6 msec at 2/sec by a mirror m o u n t e d on a galvanometer. Triangular 30 min fixation marks were attached at 30 min b e y o n d the upper and lower edges o f the target area. Because of the circular stimulus, small segments o f the hemiretina near the horizontal meridian were unstimulated. We chose this arrangement in order to reduce accidental stimulation o f the wrong halffield to a minimum in case o f slippage of fixation. Actually, differences between responses to rectangular and to circular hemiretinal stimuli were very minor.
I
R B
R-A
UPPER
HEMI-RETINA
At 4 min intervals, the subject was asked to put his head into the headrest, close one eye with a shutter attached to the headrest, and to fixate the upper or lower mark. 10 sec after he report ed 'ready' via the intercom, data collection in 39 simultaneous channels was started, and ended after 140 presentations (one run). The subject's face and eye were observed via closed circuit TV during data collection. Repeated runs yielded very similar wave forms. The example examined in this paper is the third run for upper and lower fixation with the right eye. Thirty-nine Grass gold plated cup electrodes were applied to the scalp with Grass paste, at about equal distances (4 cm) over an area between inion and about 70% o f the inion--nasion distance, and symmetrically covering 70% of the distance between the meati acustici externi. The electrode array is indicated in the inset of Fig. 2. Data from 2 anterior temporal electrodes were om i t t ed because of muscle artefacts. Using a 48-channel data handling system (constructed by J.M. Madey
R C
t!
/
j
J ~%v
/
i
0
LOWER
!
HEMI RETINA I
103
I
146 MSEC
Fig. 1. Average potential wave forms (n = 140) evoked by checkerboard reversal shown to the upper (above) and lower (below) hemiretina. Records between 3 active midline electrodes ('C' at inion, 'B' at 12%, and 'A' at 24% of inion--nasion distance) and midline reference ('R', at about 30% nasion--inion distance). Positivity of active electrodes represented by downward deflection. Zero potential difference indicated by dotted line. Times of maximal amplitudes of occipital positivity in the 2 stimulus conditions marked by vertical lines. Analysis time 256 msec, amplitude calibration 6.5 pV. From same data as Figs. 2 and 3.
HEMIRETINA-EVOKED POTENTIALS
727
and V. Corti) the EEG data were pre-amplifled (50% down at 0.5 and 100 c/sec, 12 dB per octave), sampled at 750 s/sec/channel, and multiplexed in groups of 8 channels for AM recording on 6 tracks of an instrumentation tape recorder. Off-line demultiplexing, A-D conversion and averaging in a laboratory computer were followed by calibration and transformation of the data into sequences of maps of potential distributions on the scalp in a large computer system (Lehmann 1971, 1975). Evoked potential wave forms, as conventionally recorded between electrode pairs, were reconstructed from the multichannel data stored in the large computer system (samples in Fig. 1).
Results Fig. 1 illustrates 3 conventional evoked potential wave forms reconstructed from the multichannel data, for the 2 stimulus conditions: the potentials show the weU-known approximate inversion of polarity between responses to upper and lower hemiretinal stimulation recorded between occipital and anterior electrodes. From the wave forms evoked by upper hemiretinal stimulation, the latency of the largest positive component was determined (103 msec, record R-B, upper part of Fig. 1), which more or less coincided in time with the first prominent negative peak evoked by lower hemiretinal stimulation.
UPPER HEMI-RETINA
LOWER HEMI-RETINA
88
103
118
132
146
161 MSEC
Fig. 2. Average (n = 140) scalp field distributions evoked by upper (above) and lower (below) hemiretinal stimulation with checkerboard reversal, at different times (indicated in msec) after stimulus reversal. The electrode array is schematically illustrated in the inset. The field values are computer-interpolated between electrode sites, and mapped as equipotential line plots. Each line indicates a 1 pV step. Positive field maxima indicated by plus signs, negative maxima by minus signs. Note that the location, and thus the potential value, at the c o m m o n reference electrode is o f no consequence for individual maps. The right anterior positivity at 103/118 msec for upper, and 146/161 msec for lower hemiretinal stimulation is not contingent upon the eye stimulated: it was observed with right eye and with left eye stimulation. F r o m same data as Figs. 1 and 3.
728
D. LEHMANN ET AL.
Similarly, the latency of the largest positive c o m p o n e n t evoked by lower hemiretinal stimulation (146 msec, record R-C, lower part of Fig. 1) about corresponded in time with a negative peak evoked by upper hemiretinal stimulation. Fig. 2 illustrates the scalp field distributions evoked by the 2 stimuli at the 2 peak latencies, and at additional times at 15 msec intervals immediately before, between and after the peak latencies. At 88 msec after the stimulus there is an occipital positivity in both stimulus conditions, clearly visible somewhat anterior to the inion during upper stimulation, b u t also detectable at the inion during lower stimulation. At 103 msec, the occipital positivity evoked by upper stimulation reaches its peak and becomes less prominent in subsequent maps. The map sequence evoked b y lower stimulation shows continuing build-up of the occipital positivity, which reaches its maximum at 132 and 146 msec. The maps obtained at 103 msec for upper and at 132 and 146 msec for lower stimula-
tion are strikingly similar. Thus, the fields evoked b y lower stimulation showed occipital peak positivities 43 msec later than those evoked by upper stimulation. A chronotopogram (R~mond 1968) is an efficient data display for the survey of crucial features of field distribution in map sequences. If we consider only the scalp potential values obtained in our 2 stimulus conditions from the 7 midline electrodes at 15 msec intervals, the chronotopograms shown in Fig. 3 result. Obviously, the shapes of the chronotopograms produced by upper and lower hemiretinal stimulation are very similar - - except that they are shifted in time by 43 msec. The conclusion is again that not only during, but also preceding and following the times of maximal amplitude differences in the fields {maximal 'hillyness'; Lehmann 1971), the field distributions evoked b y the 2 conditions are quite similar (see also Fig. 2), strengthening the interpretation of a delayed response evoked by lower hemiretinal stimulation.
UPPER HEMI RETINA 43
58
73
88
103
118
132
146
161 MSEC
132
146
161
175
191
205 MSEC
ANTERIOR
INION LOWER HEMI RETINA 88
103
118
ANTERIOR
INION
Fig. 3. Chronotopograms of average EEG potentials evoked by checkerboard reversal shown to upper and lower hemiretina. Illustrated are potential values along a midline row of 7 electrodes (vertical) at indicated times after stimulus reversal, in 15 msee intervals (horizontal). Pcaitivity white, negativity hatched; dotted line connects maximal positive values; solid line connects maximal negative values. Isopotential lines at 0/~V, which is mean value of the voltages recorded from all 37 electrodes at the given moment (average reference). Note similarity of c h r o n o t o p o g r a m s , b u t also shift in t i m e after stimulus, b e t w e e n t h e t w o conditions. F r o m same data as Figs. 1 and 2.
HEMIRETINA-EVOKED POTENTIALS Discussion Obviously, the approximate wave form inversion o f responses to upper and lower hemiretina stimulation in conventional records comes a b o u t by a time shift of the positive response m a x i m u m to lower stimulation which makes it coincide more or less with the subsequent negative response maxim u m to upper stimulation. Minor optical differences of light beams entering the eye at different angles cannot explain the wave form differences in the 2 conditions, since wave forms remain constant over a wide range of stimulus parameters (contrast sharpness, light intensity: Lehmann and Mir 1976). The present paper offers new interpretations based on topographical evaluations which became possible because of more extensive data collection. Our data -- as far as electrode placements are comparable - - a r e in agreement with those of other studies using similar stimuli (e.g. Michael and Halliday 1971). The data collected below the inion b y these authors confirm the positive field peaks at about the inion which our Fig. 2 indicates at 88 and 103 msec for lower hemiretinal stimulation. Our conclusion that the different wave forms reflect responses of different latencies, and not of inverted polarities, is in agreement with a report b y Eason et al. (1967) w h o found longer latencies of EEG responses evoked by xenon flashes to the lower hemiretina than to the upper hemiretina. In fact, the observed wave form inversion in the two conditions is often only very approximate, i.e. sometimes the latency differences are small (Lehmann and Mir 1976). In any case, differences between conventionally recorded EEG waves (simultaneous recordings, or different conditions) require careful evaluation, since they may be caused by the field values at the reference electrode (Lehmann 1975). An example of the influence o f the potential value at the reference electrode is the negative peak at 103 msec in our conventionally recorded wave forms evoked b y lower hemi-
729 retinal stimulation: at this time, the field value at the reference 'R' of Fig. i was somewhat more positive than the values at 'A' and 'B' (Figs. 1 and 2) at the ascending slope o f the occipital positivity. The wave form R-C of Fig. 1 (lower row) illustrates that a negative peak of a conventional wave form may actually be a near-zero difference value which is preceded and followed b y positive differences. Besides these electrophysiological data, there are classical anatomical, perceptual and behavioural findings which indicate that the human upper hemiretinal system is more powerful than its lower counterpart (see pp. 327--328 in W o o d w o r t h 1938). The upper human hemiretina has more rods and cones per mm 2 than the lower half (Osterberg 1935). There are fewer intracortical projections from lower than from upper hemiretinal cortical areas (Zeki, personal communication). Visual acuity is better for the upper hemiretina (Landolt and Hummelsheim 1904), in agreement with the assumption that denser receptor packing correlates with higher acuity (ten Doesschate 1946). In addition, m o t o r reaction time to light flashes in the upper hemiretina is shorter than in the lower (first reported by Hall and von Kries 1879). In which way the different delays of the evoked responses relate to these data will have to be clarified. At any rate, it could be argued that phylogenetically, human enemies and prey were expected in the lower half of the visual field and therefore, improved performance of the upper hemiretinal system was advantageous.
Summary Checkerboard reversals shown to the upper hemiretina evoke EEG potentials which in anterior-posterior derivations are approximately inverted in polarity compared with potentials evoked by lower hemiretinal stimulation. Using a 48-channel system, EEG scalp
730
field distributions of such responses were mapped. Examination of the map sequences shows that occipital positive maximal field values start to develop at about the same time in the two stimulus conditions, b u t peak much earlier and somewhat more anteriorly for upper than for lower hemiretinal stimulation. Thus, there is a difference of response latency in the two conditions, which accounts for the approximate inversion of polarity. Possible correlations with reports of higher receptor cell density, higher visual acuity, and shorter m o t o r reaction time of the upper hemiretinal system are noted.
Rdsumd Repdrage par canaux multiples des champs de potentiel EEG moyens obtenus d partir des demi-rgtines supgrieure et infdrieure. Diffdfences de latences Une stimulation par damier de la demir~tine sup~rieure ~voque des rdponses EEG qui, en d~rivation antdro-post~rieure, sont pratiquement de polarit~ inverse de celles ~voqu~es par stimulation de la demi-rdtine inf~rieure. La distribution des champs sur le scalp correspondent ~ ces rdponses a dtd dtablie l'aide d'un syst~me ~ 48 canaux. L'examen des cartes montre que les maxima de positivit4 commencent au m~me m o m e n t dans les deux conditions de stimulation, mais que le sommet est atteint plus t6t, et est plus ant~rieur, pour la stimulation sup~rieure que pour celle de la demi-r~tine inf~rieure. Une difference de latence de r~ponse apparaft ainsi entre les deux conditions, ce qui rend compte de l'inversion approximative des polarit~s. On associe ces r~sultats ~ ceux qui signalent une plus grande densitd de r~cepteurs, une plus grande acuit~ visuelle, dans la demi-rdtine sup~rieure, ainsi qu'un temps de r~action visuomoteur inf~rieur ~ partir de sa stimulation.
D. LEHMANN ET AL.
References Eason, R.G., White, C.T. and Oden, D. Averaged occipital responses to stimulation of sites in the upper and lower halves of the retina. Perception Psychophysics, 1967, 2 (10): 423--425. Cappin, J.M. and Nissim, S. Pattern visual evoked responses in the detection of field defects in glaucoma. Arch. Ophthal., 1975, 93: 9--18. Jeffreys, D.A. and Axford, J.G. Source locations of pattern-specific components of human visual evoked potentials. II. Component of extrastriate cortical origin. Exp. Brain Res., 1972, 16: 22--40. Hall, G.S.. und yon Kries, J. Ueber die Abh~/ngigkeit der Reaktionszeit vom Ort des Reizes. Arch. Anat. Physiol. (Leipzig), 1879, Suppl.: 1--10. Halliday, A.M. and Michael, W.F. Changes in patternevoked potentials in man associated with the vertical and horizontal meridians of the visual field. J. Physiol. (Lond.), 1970, 208: 499--513. Halliday, A.M., McDonald, W.I. and Mushin, J. Delayed visual evoked response in optic neuritis. Lancet, 1972, 1 : 982--985. Halliday, A.M., McDonald, W.I. and Mushin, J. Visual evoked response in diagnosis of multiple sclerosis. Brit. Med. J., 1973, 4: 661--664. Londolt, E. und Hummeisheim, E. Die Untersuchung der Funktionen des excentrischen Netzhautgebietes. In T. Saemisch (Ed.), Graefe-Saemisch Handbuch der Gesamten Augenheilkunde, Vol. 4. Engelmann, Leipzig, 1904: 503--583. Lehmann, D. Multichannel topography of human alpha EEG fields. Electroenceph. clin. Neurophysiol., 1971, 31: 439--449. Lehmann, D. EEG phase differences and their physiological significance in scalp field studies. In E. Dolce and H. Kiinkel (Eds.), Computerized EEG analysis. Fischer, Stuttgart, 1975 : 102--110. Lehmann, D. und Mir, Z. Methodik and Auswertung visuell evozierter EEG-Potentiale bei Verdacht auf multiple Sklerose. J. Neurol., 1976, 213: 97--103. Lehmann, D., Meles, H.P. and Mir, Z. Scalp field maps of averaged EEG potentials evoked by checkerboard inversion. Biomed. Technik, 1976, 21 (Suppl.) 117--118. Les~vre, N. Potentiels ~voqu~s par des patterns chez l'homme: influence de variables caracterisant le stimulus et sa position dans le champ visuel. In A. Fessard et G. Lelord (Eds.), Activit~s ~voqu~es et leur conditionnement. INSERM, Paris, 1973: 1--22. Michael, W.F. and Halliday, A.M. Differences between the occipital distribution of upper and lower field pattern evoked responses in man. Brain Res., 1971, 32: 311--324. ¢}sterberg, G. Topography of the layer of rods and
HEMIRETINA-EVOKED POTENTIALS cones in the human retina. Acta ophthal. (Kbh.), 1935, Suppl. 6: 1--102. Rdmond, A. The importance of topographic data in EEG phenomena, and an electrical model to reproduce them. Electroenceph. clin. Neurophysi'ol., 1958, Suppl. 27: 27--49. Spekreijse, H., van der Tweel, L.H. and Zuidema, T.
731
Contrast evoked responses in man. Vision Res., 1973, 13: 1577--1601. Ten Doesscha~, J. Visual acuity and distribution of percipient elements on the retina. Ophthalmologica (Basel), 1946, 112: 1--18. Woodworth, R.S. Experimental psychology. Holt, N e w York 1938.
