- Email: [email protected]
Further investigation of visual evoked potentials elicited by lateralized stimuli: Effects of stimulus eccentricity and reference site
Electroencephalography and clinical Neurophysiologv, 1985, 62:81-87 Elsevier Scientific Publishers Ireland, Ltd.
81
F U R T H E R INVESTIGATION O F VISUAL EVOKED P O T E N T I A L S ELICITED BY LATERALIZED STIMULI: EFFECTS O F S T I M U L U S ECCENTRICITY AND REFERENCE SITE MICHAEL D. RUGG, CHRISTOPHER R. LINES and A. DAVID MILNER
MRC Cognttive Neuroscience Research Group, Psychological Laboratory, University of St. Andrews, St. Andrews, Fife. KY16 9J U (U.K.) (Accepted for publication: September 7, 1984)
Attempts to use visual evoked potentials (VEPs) to study interhemispheric transfer in man have been reported sporadically over the past 15 years or so (see Rugg 1982 for a review). Recently, we have investigated in some detail VEPs to small lateralized light flashes in an attempt to determine whether such a technique may provide a reliable means of indexing the interhemispheric transfer of visual information (Lines et al. 1984; Rugg et al. 1984a,b). The principal results of these investigations are: (i) A prominent relatively early negative VEP component, N160, peaks earlier and is of a larger amplitude when recorded from over the hemisphere contralateral to the visual field of stimulus exposure. (ii) These N160 latency asymmetries are greater between homotopic occipital electrodes (typically 10-15 msec) than between central electrodes (typically 3-4 msec). (iii) These ipsilateral-contralateral latency differences increase as stimulus intensity is decreased at occipital electrode sites, but not at central sites (Lines et al. 1984). (iv) Two patients with congenital absence of the corpus callosum show normal contralateral VEPs, but lack the ipsilateral N160 component seen in normals (Rugg et al. 1984b). We have drawn two main conclusions from these studies. Firstly, that the ipsilateral N160 depends for its generation on transcallosal processes and secondly, that occipital and central homotopic pairs of electrodes are detecting transcallosal relay in at least two functionally and anatomically distinct regions of the corpus callo-
sum, the properties of which are consistent with the model of information flow between and within the hemispheres proposed by Milner and Lines (1982). The present study extends the above findings in two ways. We compare VEPs elicited by stimuli at the lateral eccentricity of 4 ° , as employed in all our previous work, with those elicited by stimuli at an eccentricity of 10 ° . The difference between these eccentricities is relatively small in terms of changes in probable numbers of retinotopic transcallosal connections between areas of visual cortex (Berlucchi 1981; Lepore and Guillemot 1982). It allows an assessment of the possibifity that VEPs recorded over the hemisphere ipsilateral to the visual field of stimulus exposure may be subject to contamination by light scattering from the stimulated hemi-field across the visual midline, into the supposedly unstimulated hemi-field. If this is the case, then ipsilateral-contralateral asymmetries should be modified by the 6 ° increase in eccentricity investigated in this study, as this should confine far more of any scattered light to the stimulated visual field. A second aim of this study was to determine the extent to which the effects we have reported previously, particularly the dissociations in N160 latencies between central and occipital sites, depend upon the employment of linked mastoids as a reference. This reference has been shown to be unsuitable for the study of the lateral distributic~n of VEPs generated by pattern-onset or reversing checkerboard stimuli because of the extensive lateral spread of the resulting VEP fields (Halliday et al. 1977). Although the VEPs elicited by the
0168-5597/85/$03.30 '~ 1985 Elsevier Scientific Publishers Ireland, Ltd.
82 small flashes used in our studies are smaller and appear to be less extensively distributed over the scalp than pattern-generated VEPs, the comparison of the linked mastoids reference with a potentially more 'indifferent' reference site is an important methodological requirement.
Methods
Subjects These consisted of 14 right-handed young adults, 10 of whom were male. All had normal or corrected vision. Stimuli and task Stimuli consisted of flashes emitted by TTLdriven LEDs (RS Components, Catalogue number 588-285) situated either 4 ° or 10 ° lateral to fixation. They subtended a visual angle of 0.33 °, and when illuminated emitted light with an intensity of 140 mcd at a wave length of 585 nm. Two exposure durations were employed: G O stimuli were exposed for 5 msec and N O - G O stimuli for 40 msec. To control for any differences in stimulus characteristics the LEDs were interchanged halfway through the experimental runs at each eccentricity. A G O / N O - G O choice reaction time task was employed. This required subj-~ts to make a finger push to G O stimuli as fast as possible with the hand ipsilateral to the visual field of stimulus occurrence. In each experimental run a total of 100 stimuli were exposed randomly in the left and right visual fields with an inter-stimulus interval of 2 sec. The probability of occurrence of a G O stimulus was 0.24 and a warning tone preceded the onset of each stimulus by 600 msec. This task has two advantages over one in which stimuli are observed passively: (i) it encourages a reasonable level of alertness and minimizes contamination of VEPs with alpha activity, and (ii) it helps prevent, and allows detection of, systematic attentional biases to one visual field. By using a relatively high proportion of N O - G O trials, these advantages are achieved while allowing the collection of VEPs uncontaminated by motor responses.
M.D. RUGG ET AL.
Procedure Subjects viewed a dim, constantly illuminated fixation light in a darkened room, with their heads restrained by a chin rest and the index finger of each hand resting on laterally positioned microswitches. They were given 100 practice trials and a further 8 blocks of 100 trials each. The first four of these blocks were with the stimuli at one of the two eccentricities, and the remainder with stimuli at the other, the starting eccentricity being counterbalanced over subjects. A total of 152 N O - G O trials were thus exposed in each visual field in each condition. Subjects were informed of the importance of maintaining fixation and minimizing eye and body movements. They were allowed to rest for as long as they wished between blocks./ VEP recording E E G was recorded with silver/silver chloride electrodes postioned at Pz, C3, C4 and left and right mastoid processes. The occipital electrodes, designated here as LO1 and LO2, were somewhat lateral to the standard O1 and 0 2 sites, 20% of the distance along the line from Oz to Fpz on the left and right. All these electrodes were referred to the balanced non-cephalic reference described by Stephenson and Gibbs (1951). The bandpass of these channels was 0.16-100 Hz (3 dB points). E O G was recorded bipolarly from electrodes situated on the outer canthus of the left eye and just above the outer canthus of the right eye. The bandpass of this channel was 0.03-30 Hz. All channels were digitized and averaged on-line. Sampling began 60 msec before stimulus exposure and continued for 708 msec thereafter. To ensure that VEPs were uncontaminated by eye movement artefact, and that they did not include trials on which subjects made a saccade towards the exposed stimulus, a criterion level of E O G amplitude was established for each subject prior to the experimental runs. This was determined so that any E E G epochs associated with blinks or saccades were automatically rejected, and E O G was averaged along with the E E G so that the success of this procedure could be verified for each subject. In addition, epochs associated with incorrect behavioural responses (errors of omission or commission) were also excluded from the averages.
VEPs E L I C I T E D BY L A T E R A L I Z E D S T I M U L I
83
particularly at occipital sites. Similar, though much attenuated and rather ill-defined, negative deflections can be observed at the contralateral and ipsilateral mastoid sites. At the Pz and occipital sites only, an earlier P120 component is also present and is particularly prominent at the ipsilateral occipital electrodes. Qualitatively, the data from the scalp electrodes closely resemble those previously obtained with a linked mastoid reference (e.g., Lines et al. 1984; Rugg et al. 1984a). Quantification of the VEPs from scalp sites was achieved by visually guided cursor measurement. This procedure was supplemented by an objective cross-
Results
Grand average VEPs at each electrode site and in each condition are shown in Fig. 1. These and subsequent analyses are based on only 12 of the 14 subjects; in the remaining two (both males) it proved impossible to eliminate ECG artefact from the chest electrodes of the non-cephalic reference. At all scalp electrodes the prominent N160 component can be observed and, as in previous studies, it appears to peak earlier and with an increased amplitude over the hemisphere contralateral to the visual field of stimulus exposure,
4°
10 °
LVF
RVF
LVF
RVF
Pz Nt60
Central
.,'-.
Lateral ,.':...~ occipital ~'-'--~J
Mastoid ~ . . _ ~ ~
....,,,~~
I
0
300
0
0
300
-
-
i
1
l
1
I
i
300
Contralateral Ipsilateral
t
0
electrode
Lateral s i t e s : ................
J
electrode
Fig. l. Grand average VEPs at each electrode site and eccentricity using a non-cephalic reference.
I
I
1
--
300
I
1
L--I
msec
84
M , D . R U G G E T AL.
correlation procedure (described in Lines et al. 1984) which assessed ipsilateral-contralateral latency differences for each homotopic pair of lateral electrodes for each visual field. The measures obtained by these two procedures were compared and any major discrepancies resolved. N160 was measured at all electrode sites in all subjects. P120 was measured from Pz and occipital sites only, as it was small and ill-defined at the central sites in the majority of subjects. Peak latencies and amplitudes (measured with respect to the mean amplitude of the pre-stimulus baseline) were assessed with repeated measures ANOVAs, performed separately on the data from the midline and lateral electrodes.
Effects of stimulus eccentricity Stimulus eccentricity had no significant effect on the latencies of P120 or N160 at either the midline or lateral sites and these data are shown collapsed over eccentricity in Table I. Analysis of P120 latency at the occipital sites revealed only a significant interaction between visual field and electrode site (F1.11 = 22.13, P < 0.001), reflecting the fact that the component from the contralateral electrode peaked on average 14 msec earlier than its ipsilateral counterpart. Analysis of N160 latency recorded from the lateral electrodes revealed a significant main effect for electrode site ( F 3 , 3 3 = 22.13, P < 0 . 0 0 0 1 ) and an interaction between electrode site and visual field ( F 3 , 3 3 = 14.93, P < 0.0001). The former effect is due to this component peaking on average 15 msec earlier at the central sites than at the occipital electrodes. The interaction is caused by N160 latency being shorter when recorded from contralateral electrodes. Posthoc analyses (Scheffe's procedure) showed that
while this contralateral latency advantage was reliable at the occipital sites, where it averaged 10 msec, it fell far short of significance at the central sites, where a considerably smaller mean value of 2 msec was obtained. Analysis of P120 amplitude at Pz and occipital sites also revealed no evidence of an eccentricity effect. A significant interaction between visual field and electrode site was found at the occipital sites ( F 1 , 1 1 = 14.26, P < 0.005), indicating that the amplitude of this component was greater at each electrode when ipsilateral to the visual field of stimulus exposure; this ipsilateral-contralateral difference averaged 2.1/~V. No effect of eccentricity was observed for N160 amplitude at the Pz electrode. However, at the lateral sites a significant visual field by electrode site interaction was revealed ( F 3 , 3 3 = 18.40, P < 0.0001). Post-hoc analysis showed that was caused by the contralateral electrode giving rise to greater N160 amplitude at both central and occipital sites, with contralateral advantages of 1.6 #V and 3.3/zV respectively. This interaction was accompanied by a significant second-order interaction between visual field, electrode site and eccentricity ( F 3 , 3 3 = 5 . 7 0 , P < 0.005). This reflected a tendency for the above contralateral N160 amplitude advantage at the occipital sites to be reduced at 10 ° eccentricity with right visual field stimuli.
Non-cephalic vs. mastoid reference Difference wave forms from the 4 ° condition were formed for each subject by subtracting the appropriate averaged left and right mastoid activity from the VEPs recorded with the non-cephalic reference, thus 'reconstructing' the VEPs as they would have appeared if originally recorded against
TABLE I P120 a n d N 1 6 0 latencies (msec) b y electrode site a n d visual field (collapsed across eccentricities). LVF Pz
RVF C3
LO1
P120
Mean S.D.
118 14
-
128 9
N160
Mean S.D.
159 9
154 7
174 12
C4 154 6
LO2
Pz
118 16
117 16
166 12
157 10
C3 152 6
LO1 113 14 163 11
C4 156 7
LO2 130 8 174 12
85
VEPs ELICITED BY LATERALIZED STIMULI LVF
(F1,11 = 26.39, P < 0.0005). At the lateral sites, a
RVF
similar overall reduction in amplitude of 1.4 /~V was found (Fl, la =42.08, P < 0.0001). A significant interaction between type of reference and electrode site (F3,33 = 5.08, P < 0.01) indicated that this reduction was greater at the occipital (mean 1.8 /tV) than the central sites (mean = 1.1
Pz +
Central
~v).
lOuV
Discussion
Lateral occipital
I
0
l
l
t
l
l
[
I
i
300
0
-
-
I
I
l
I
I
I
i
300
msec
Contralateral electrode
Lateral sites: .............
Ipsilate'tal e l e c t r o d e
Fig. 2. Grand average VEPs at each electrode site to stimuli at 4° using a linked mastoid reference.
linked mastoids. The grand averages of these VEPs are shown in Fig. 2. PI20 and N160 were quantified using the procedures described previously, and these data were compared with those obtained with the non-cephalic reference from the 4 ° condition of the experiment, using repeated measures ANOVA. Only those effects indicating an influence of reference site, or those indicating a difference from the pattern of results obtained with the inclusion of the 10 ° data in the previous analyses, will be described. Analysis of P120 latency and amplitude revealed no influence of reference site on the magnitude or patterning of these data. However, N160 latency was affected by reference site, although only at the lateral electrodes (main effect of type of reference: F1, 11 = 10.28, P < 0.01). This effect did not interact with any other factor and was caused by the fact that N160 latencies peaked on average 4 msec earlier in the VEPs recorded against the mastoid reference. Type of reference also influenced N160 amplitude. Its magnitude at Pz was reduced by 1.3 ~V in the wave forms recorded against the mastoids
The data from this study confirm our previous findings with respect to N160 (Lines et al. 1984; Rugg et al. 1984a) and demonstrate the same pattern of results for P120, at least at occipital electrodes. VEPs were almost entirely unaffected by the stimulus eccentricity manipulation employed in this experiment, particularly with regard to ipsilateral-contralateral differences in the peak latencies of P120 and N160. This lends further weight to the claim (originally made for other reasons by Lines et al. 1984) that the ipsilateral VEPs observed in this paradigm are not significantly affected by the scattering of light across the visual midline. It remains to be seen whether the employment of a wider eccentricity range, thus allowing the comparison of interhemispheric transfer occurring via substantially different numbers of callosal fibres, would result in detectable differences in ipsilateral-contralateral latency asymmetries. Such studies, were they to include eccentricities nearer the visual midline, may help to determine the level in the visual system at which this transfer occurs. Extrapolation from neurophysiological studies in the monkey (Van Essen et al. 1982) suggests that few callosal fibres terminate in areas of striate or pre-striate cortex subserving the eccentricities employed in the current study. However, such fibres do project extensively to visual areas of the temporal lobe (Gross et al. 1977; Bruce et al. 1981). It is possible, therefore, that the ipsilateral VEP recorded from occipital electrodes in our paradigm relies for its generation on visual information transmitted to the human homologue of these visual areas of monkey temporal lobe. The other notable feature of these results is that
86
in comparison with a non-cephalic reference, the use of linked mastoids had no effect on P120 latency and amplitude and only non-specific effects on these measures of the N160 component. Thus, our previous findings that N160 peaks earlier at central compared with occipital sites, and shows larger and more reliable ipsilateral-contralateral latency differences at the latter sites, are unmodified; the same is true with respect to visual field dependent N160 amplitude asymmetries. The somewhat larger differences in N160 amplitude observed at occipital sites in the comparison of the two references is accounted for by the fact that the stimulus-locked negative deflections occurring at the mastoids peak at approximately the same latency as the occipital N160s, and thus somewhat earlier than those recorded from the central sites (Fig. 1). To conclude, we have shown that previously reported ipsilateral-contralateral asymmetries in the amplitudes and latencies of two VEP components elicited by small lateralized light flashes are unlikely to result either from light scatter or an inappropriate choice of reference site. To the extent that these data are considered informative about transcallosal processes (our reasons for considering them as such are elucidated in Lines et al. 1984 and Rugg et al. 1984a) this technique may provide a useful means for their assessment.
Summaff Visual evoked potentials (VEPs) to small lateralized flashes were recorded from the parietal midline, homotopic lateral central and occipital electrodes, and from left and right mastoid processes, all referred to a balanced non-cephalic reference. Two stimulus eccentricities, 4 ° and 10 °, were employed, and a comparison made between the non-cephalic and linked mastoid references. P120 (measured at lateral occipital sites only) peaked earlier and was of smaller amplitude at the electrode contralateral to the visual field of stimulus exposure. N160 peaked earlier at central than occipital sites, was larger from electrodes over the contralateral hemisphere, and at the occipital sites only, peaked earlier in the electrode contralateral
M,D. R U G G ET AL.
to the visual field of stimulus exposure. These effects were virtually unaffected by the eccentricity manipulation and it is concluded that light scatter across the visual midline is unlikely to be responsible for the observed pattern of ipsilateral-contralateral VEP asymmetries. The mastoids were found to detect significant stimulus-locked activity in the same latency range as the occipital N160 component. However, comparison of the non-cephalic and linked mastoid references revealed only non-specific effects, and no sign of any change in the pattern of ipsilateralcontralateral VEP asymmetries, or the magnitude of the associated latency differences. It is concluded that these effects may be of value in the study of callosal transfer.
R6sum6 Nouvelle &ude des potentiels bvoqubs v&uels par des stimulus latbralisks: effets de l'excentricitb du stimulus et du site de rbfbrence Des potentiels visuels 6voqurs (PEV) par des flash 16grrement latrralisrs furent enregistrrs par des 61ectrodes situres en parirtal mrdian, en centrale latrrale homotopique et occipitale, et sur les mastoides, toutes avec une rbfrrence non crphalique compensre. Deux excentricitrs du stimulus, de 4 ° et 10 °, furent utilisres et une comparaison entre les rrfrrences non c6phalique et mastoidienne rrunie a 6t6 faite. La P120 (mesurre aux sites occipitaux seulement) prrsentait un pic plus prrcoce et de plus faible amplitude h l'61ectrode contralatrrale au champ visuel expos6 au stimulus. N160 avait un pic plus prrcoce au sites centraux qu'occipitaux, 6tait plus grand pour les 61ectrodes sur l'hrmisph~re contralatrral au champ visuel soumis au stimulus. Ces effets n'rtaient virtuellement pas affectrs par la modification de l'excentricit6 et on en conclut qu'une dispersion de la lurnirre autour de la ligne visuelle mrdiane n'est vraisemblablement pas responsable du pattern observ6 d'asymrtrie des PEV ipsi- et contralatrraux. Les mastoides permettent de drtecter une activit6 significative li6e au stimulus dans le mrme
VEPs ELICITED BY LATERALIZED STIMULI d o m a i n e d e l a t e n c e q u e la c o m p o s a n t e o c c i p i t a l e N 1 6 0 . T o u t e f o i s , la c o m p a r a i s o n d e s r r f r r e n c e s n o n c r p h a l i q u e s et m a s t o ' i d i e n n e s r r u n i e s n ' a r r v r 1 6 que des effets non sprcifiques, et aucun signe de m o d i f i c a t i o n d a n s le p a t t e r n d ' a s y m r t r i e d e s P E V i p s i - et c o n t r a l a t r r a u x o u d a n s la g r a n d e u r d e s diffrrences de latences associres. On en conclut q u e ces e f f e t s p e u v e n t d t r e u t i l e s d a n s l ' r t u d e d e s transferts callosaux. This research is supported by the Wellcome Trust and the Medical Research Council of the U.K.
References Berlucchi, G. Recent advances in the analysis of the neural substrates of interhemispheric communication. In: O. Pompeiano and C. Ajmone Marsan (Eds.), Brain Mechanisms and Perceptual Awareness. Raven Press, New York, 1981: 133-152. Bruce, C.J., Desimone, R. and Gross, C.G. Visual properties of neurons in a polysensory area in superior temporal sulcus of the macaque. J. Neurophysiol., 1981, 46: 369-384. Gross, C.G., Bender, D.B. and Mishkin, M. Contributions of the corpus callosum and the anterior commissure to visual activation of inferior temporal neurons. Brain Res., 1977, 131: 227-239.
87 Halliday, A.M., Barrett, G., Halliday, E. and Michael, W.F. The topography of the pattern evoked potential. In: J.E. Desmedt (Ed.), Visual Evoked Potentials in Man: New Developments. Clarendon, Oxford, 1977: 121-133. Lepore, F. and Guillemot, J.-P. Visual receptive field properties of cells innervated through the corpus callosum in the cat. Exp. Brain Res., 1982, 46: 413-424. Lines, C.R., Rugg, M.D. and Milner, A.D. The effect of stimulus intensity on visual evoked potential estimates of interhemispheric transmission time. Exp. Brain Res., 1984, in press. Milner, A.D. and Lines, C.R. Interhemispheric pathways in simple reaction time to lateralized light flash. Neuropsychologia, 1982, 20: 171-179. Rugg, M.D. Electrophysiological studies. In: J.G. Beaumont (Ed.), Divided Visual Field Studies of Cerebral Organisation. Academic Press, London, 1982: 129-145. Rugg, M.D., Lines, C.R. and Milner, A.D. Visual evoked potentials to lateralised visual stimuli and the measurement of interhemispheric transmission time. Neuropsychologia, 1984a, 22: 215-225. Rugg, M.D., Milner, A.D. and Lines, C.R. Visual evoked potentials to lateralised stimuli in two cases of callosal agenesis. J. Neurol. Neurosurg. Psychiat., 1984b, in press. Stephenson, W.A. and Gibbs, F.A. A balanced non-cephalic reference electrode. Electroenceph. clin. Neurophysiol., 1951, 3: 237-240. Van Essen, D.C., Newsome, W.T. and Bixby, J.L. The pattern of interhemispheric connections and its relationship to extrastriate visual areas in the macaque monkey. J. Neurosci., 1982, 2: 265-283.
81
F U R T H E R INVESTIGATION O F VISUAL EVOKED P O T E N T I A L S ELICITED BY LATERALIZED STIMULI: EFFECTS O F S T I M U L U S ECCENTRICITY AND REFERENCE SITE MICHAEL D. RUGG, CHRISTOPHER R. LINES and A. DAVID MILNER
MRC Cognttive Neuroscience Research Group, Psychological Laboratory, University of St. Andrews, St. Andrews, Fife. KY16 9J U (U.K.) (Accepted for publication: September 7, 1984)
Attempts to use visual evoked potentials (VEPs) to study interhemispheric transfer in man have been reported sporadically over the past 15 years or so (see Rugg 1982 for a review). Recently, we have investigated in some detail VEPs to small lateralized light flashes in an attempt to determine whether such a technique may provide a reliable means of indexing the interhemispheric transfer of visual information (Lines et al. 1984; Rugg et al. 1984a,b). The principal results of these investigations are: (i) A prominent relatively early negative VEP component, N160, peaks earlier and is of a larger amplitude when recorded from over the hemisphere contralateral to the visual field of stimulus exposure. (ii) These N160 latency asymmetries are greater between homotopic occipital electrodes (typically 10-15 msec) than between central electrodes (typically 3-4 msec). (iii) These ipsilateral-contralateral latency differences increase as stimulus intensity is decreased at occipital electrode sites, but not at central sites (Lines et al. 1984). (iv) Two patients with congenital absence of the corpus callosum show normal contralateral VEPs, but lack the ipsilateral N160 component seen in normals (Rugg et al. 1984b). We have drawn two main conclusions from these studies. Firstly, that the ipsilateral N160 depends for its generation on transcallosal processes and secondly, that occipital and central homotopic pairs of electrodes are detecting transcallosal relay in at least two functionally and anatomically distinct regions of the corpus callo-
sum, the properties of which are consistent with the model of information flow between and within the hemispheres proposed by Milner and Lines (1982). The present study extends the above findings in two ways. We compare VEPs elicited by stimuli at the lateral eccentricity of 4 ° , as employed in all our previous work, with those elicited by stimuli at an eccentricity of 10 ° . The difference between these eccentricities is relatively small in terms of changes in probable numbers of retinotopic transcallosal connections between areas of visual cortex (Berlucchi 1981; Lepore and Guillemot 1982). It allows an assessment of the possibifity that VEPs recorded over the hemisphere ipsilateral to the visual field of stimulus exposure may be subject to contamination by light scattering from the stimulated hemi-field across the visual midline, into the supposedly unstimulated hemi-field. If this is the case, then ipsilateral-contralateral asymmetries should be modified by the 6 ° increase in eccentricity investigated in this study, as this should confine far more of any scattered light to the stimulated visual field. A second aim of this study was to determine the extent to which the effects we have reported previously, particularly the dissociations in N160 latencies between central and occipital sites, depend upon the employment of linked mastoids as a reference. This reference has been shown to be unsuitable for the study of the lateral distributic~n of VEPs generated by pattern-onset or reversing checkerboard stimuli because of the extensive lateral spread of the resulting VEP fields (Halliday et al. 1977). Although the VEPs elicited by the
0168-5597/85/$03.30 '~ 1985 Elsevier Scientific Publishers Ireland, Ltd.
82 small flashes used in our studies are smaller and appear to be less extensively distributed over the scalp than pattern-generated VEPs, the comparison of the linked mastoids reference with a potentially more 'indifferent' reference site is an important methodological requirement.
Methods
Subjects These consisted of 14 right-handed young adults, 10 of whom were male. All had normal or corrected vision. Stimuli and task Stimuli consisted of flashes emitted by TTLdriven LEDs (RS Components, Catalogue number 588-285) situated either 4 ° or 10 ° lateral to fixation. They subtended a visual angle of 0.33 °, and when illuminated emitted light with an intensity of 140 mcd at a wave length of 585 nm. Two exposure durations were employed: G O stimuli were exposed for 5 msec and N O - G O stimuli for 40 msec. To control for any differences in stimulus characteristics the LEDs were interchanged halfway through the experimental runs at each eccentricity. A G O / N O - G O choice reaction time task was employed. This required subj-~ts to make a finger push to G O stimuli as fast as possible with the hand ipsilateral to the visual field of stimulus occurrence. In each experimental run a total of 100 stimuli were exposed randomly in the left and right visual fields with an inter-stimulus interval of 2 sec. The probability of occurrence of a G O stimulus was 0.24 and a warning tone preceded the onset of each stimulus by 600 msec. This task has two advantages over one in which stimuli are observed passively: (i) it encourages a reasonable level of alertness and minimizes contamination of VEPs with alpha activity, and (ii) it helps prevent, and allows detection of, systematic attentional biases to one visual field. By using a relatively high proportion of N O - G O trials, these advantages are achieved while allowing the collection of VEPs uncontaminated by motor responses.
M.D. RUGG ET AL.
Procedure Subjects viewed a dim, constantly illuminated fixation light in a darkened room, with their heads restrained by a chin rest and the index finger of each hand resting on laterally positioned microswitches. They were given 100 practice trials and a further 8 blocks of 100 trials each. The first four of these blocks were with the stimuli at one of the two eccentricities, and the remainder with stimuli at the other, the starting eccentricity being counterbalanced over subjects. A total of 152 N O - G O trials were thus exposed in each visual field in each condition. Subjects were informed of the importance of maintaining fixation and minimizing eye and body movements. They were allowed to rest for as long as they wished between blocks./ VEP recording E E G was recorded with silver/silver chloride electrodes postioned at Pz, C3, C4 and left and right mastoid processes. The occipital electrodes, designated here as LO1 and LO2, were somewhat lateral to the standard O1 and 0 2 sites, 20% of the distance along the line from Oz to Fpz on the left and right. All these electrodes were referred to the balanced non-cephalic reference described by Stephenson and Gibbs (1951). The bandpass of these channels was 0.16-100 Hz (3 dB points). E O G was recorded bipolarly from electrodes situated on the outer canthus of the left eye and just above the outer canthus of the right eye. The bandpass of this channel was 0.03-30 Hz. All channels were digitized and averaged on-line. Sampling began 60 msec before stimulus exposure and continued for 708 msec thereafter. To ensure that VEPs were uncontaminated by eye movement artefact, and that they did not include trials on which subjects made a saccade towards the exposed stimulus, a criterion level of E O G amplitude was established for each subject prior to the experimental runs. This was determined so that any E E G epochs associated with blinks or saccades were automatically rejected, and E O G was averaged along with the E E G so that the success of this procedure could be verified for each subject. In addition, epochs associated with incorrect behavioural responses (errors of omission or commission) were also excluded from the averages.
VEPs E L I C I T E D BY L A T E R A L I Z E D S T I M U L I
83
particularly at occipital sites. Similar, though much attenuated and rather ill-defined, negative deflections can be observed at the contralateral and ipsilateral mastoid sites. At the Pz and occipital sites only, an earlier P120 component is also present and is particularly prominent at the ipsilateral occipital electrodes. Qualitatively, the data from the scalp electrodes closely resemble those previously obtained with a linked mastoid reference (e.g., Lines et al. 1984; Rugg et al. 1984a). Quantification of the VEPs from scalp sites was achieved by visually guided cursor measurement. This procedure was supplemented by an objective cross-
Results
Grand average VEPs at each electrode site and in each condition are shown in Fig. 1. These and subsequent analyses are based on only 12 of the 14 subjects; in the remaining two (both males) it proved impossible to eliminate ECG artefact from the chest electrodes of the non-cephalic reference. At all scalp electrodes the prominent N160 component can be observed and, as in previous studies, it appears to peak earlier and with an increased amplitude over the hemisphere contralateral to the visual field of stimulus exposure,
4°
10 °
LVF
RVF
LVF
RVF
Pz Nt60
Central
.,'-.
Lateral ,.':...~ occipital ~'-'--~J
Mastoid ~ . . _ ~ ~
....,,,~~
I
0
300
0
0
300
-
-
i
1
l
1
I
i
300
Contralateral Ipsilateral
t
0
electrode
Lateral s i t e s : ................
J
electrode
Fig. l. Grand average VEPs at each electrode site and eccentricity using a non-cephalic reference.
I
I
1
--
300
I
1
L--I
msec
84
M , D . R U G G E T AL.
correlation procedure (described in Lines et al. 1984) which assessed ipsilateral-contralateral latency differences for each homotopic pair of lateral electrodes for each visual field. The measures obtained by these two procedures were compared and any major discrepancies resolved. N160 was measured at all electrode sites in all subjects. P120 was measured from Pz and occipital sites only, as it was small and ill-defined at the central sites in the majority of subjects. Peak latencies and amplitudes (measured with respect to the mean amplitude of the pre-stimulus baseline) were assessed with repeated measures ANOVAs, performed separately on the data from the midline and lateral electrodes.
Effects of stimulus eccentricity Stimulus eccentricity had no significant effect on the latencies of P120 or N160 at either the midline or lateral sites and these data are shown collapsed over eccentricity in Table I. Analysis of P120 latency at the occipital sites revealed only a significant interaction between visual field and electrode site (F1.11 = 22.13, P < 0.001), reflecting the fact that the component from the contralateral electrode peaked on average 14 msec earlier than its ipsilateral counterpart. Analysis of N160 latency recorded from the lateral electrodes revealed a significant main effect for electrode site ( F 3 , 3 3 = 22.13, P < 0 . 0 0 0 1 ) and an interaction between electrode site and visual field ( F 3 , 3 3 = 14.93, P < 0.0001). The former effect is due to this component peaking on average 15 msec earlier at the central sites than at the occipital electrodes. The interaction is caused by N160 latency being shorter when recorded from contralateral electrodes. Posthoc analyses (Scheffe's procedure) showed that
while this contralateral latency advantage was reliable at the occipital sites, where it averaged 10 msec, it fell far short of significance at the central sites, where a considerably smaller mean value of 2 msec was obtained. Analysis of P120 amplitude at Pz and occipital sites also revealed no evidence of an eccentricity effect. A significant interaction between visual field and electrode site was found at the occipital sites ( F 1 , 1 1 = 14.26, P < 0.005), indicating that the amplitude of this component was greater at each electrode when ipsilateral to the visual field of stimulus exposure; this ipsilateral-contralateral difference averaged 2.1/~V. No effect of eccentricity was observed for N160 amplitude at the Pz electrode. However, at the lateral sites a significant visual field by electrode site interaction was revealed ( F 3 , 3 3 = 18.40, P < 0.0001). Post-hoc analysis showed that was caused by the contralateral electrode giving rise to greater N160 amplitude at both central and occipital sites, with contralateral advantages of 1.6 #V and 3.3/zV respectively. This interaction was accompanied by a significant second-order interaction between visual field, electrode site and eccentricity ( F 3 , 3 3 = 5 . 7 0 , P < 0.005). This reflected a tendency for the above contralateral N160 amplitude advantage at the occipital sites to be reduced at 10 ° eccentricity with right visual field stimuli.
Non-cephalic vs. mastoid reference Difference wave forms from the 4 ° condition were formed for each subject by subtracting the appropriate averaged left and right mastoid activity from the VEPs recorded with the non-cephalic reference, thus 'reconstructing' the VEPs as they would have appeared if originally recorded against
TABLE I P120 a n d N 1 6 0 latencies (msec) b y electrode site a n d visual field (collapsed across eccentricities). LVF Pz
RVF C3
LO1
P120
Mean S.D.
118 14
-
128 9
N160
Mean S.D.
159 9
154 7
174 12
C4 154 6
LO2
Pz
118 16
117 16
166 12
157 10
C3 152 6
LO1 113 14 163 11
C4 156 7
LO2 130 8 174 12
85
VEPs ELICITED BY LATERALIZED STIMULI LVF
(F1,11 = 26.39, P < 0.0005). At the lateral sites, a
RVF
similar overall reduction in amplitude of 1.4 /~V was found (Fl, la =42.08, P < 0.0001). A significant interaction between type of reference and electrode site (F3,33 = 5.08, P < 0.01) indicated that this reduction was greater at the occipital (mean 1.8 /tV) than the central sites (mean = 1.1
Pz +
Central
~v).
lOuV
Discussion
Lateral occipital
I
0
l
l
t
l
l
[
I
i
300
0
-
-
I
I
l
I
I
I
i
300
msec
Contralateral electrode
Lateral sites: .............
Ipsilate'tal e l e c t r o d e
Fig. 2. Grand average VEPs at each electrode site to stimuli at 4° using a linked mastoid reference.
linked mastoids. The grand averages of these VEPs are shown in Fig. 2. PI20 and N160 were quantified using the procedures described previously, and these data were compared with those obtained with the non-cephalic reference from the 4 ° condition of the experiment, using repeated measures ANOVA. Only those effects indicating an influence of reference site, or those indicating a difference from the pattern of results obtained with the inclusion of the 10 ° data in the previous analyses, will be described. Analysis of P120 latency and amplitude revealed no influence of reference site on the magnitude or patterning of these data. However, N160 latency was affected by reference site, although only at the lateral electrodes (main effect of type of reference: F1, 11 = 10.28, P < 0.01). This effect did not interact with any other factor and was caused by the fact that N160 latencies peaked on average 4 msec earlier in the VEPs recorded against the mastoid reference. Type of reference also influenced N160 amplitude. Its magnitude at Pz was reduced by 1.3 ~V in the wave forms recorded against the mastoids
The data from this study confirm our previous findings with respect to N160 (Lines et al. 1984; Rugg et al. 1984a) and demonstrate the same pattern of results for P120, at least at occipital electrodes. VEPs were almost entirely unaffected by the stimulus eccentricity manipulation employed in this experiment, particularly with regard to ipsilateral-contralateral differences in the peak latencies of P120 and N160. This lends further weight to the claim (originally made for other reasons by Lines et al. 1984) that the ipsilateral VEPs observed in this paradigm are not significantly affected by the scattering of light across the visual midline. It remains to be seen whether the employment of a wider eccentricity range, thus allowing the comparison of interhemispheric transfer occurring via substantially different numbers of callosal fibres, would result in detectable differences in ipsilateral-contralateral latency asymmetries. Such studies, were they to include eccentricities nearer the visual midline, may help to determine the level in the visual system at which this transfer occurs. Extrapolation from neurophysiological studies in the monkey (Van Essen et al. 1982) suggests that few callosal fibres terminate in areas of striate or pre-striate cortex subserving the eccentricities employed in the current study. However, such fibres do project extensively to visual areas of the temporal lobe (Gross et al. 1977; Bruce et al. 1981). It is possible, therefore, that the ipsilateral VEP recorded from occipital electrodes in our paradigm relies for its generation on visual information transmitted to the human homologue of these visual areas of monkey temporal lobe. The other notable feature of these results is that
86
in comparison with a non-cephalic reference, the use of linked mastoids had no effect on P120 latency and amplitude and only non-specific effects on these measures of the N160 component. Thus, our previous findings that N160 peaks earlier at central compared with occipital sites, and shows larger and more reliable ipsilateral-contralateral latency differences at the latter sites, are unmodified; the same is true with respect to visual field dependent N160 amplitude asymmetries. The somewhat larger differences in N160 amplitude observed at occipital sites in the comparison of the two references is accounted for by the fact that the stimulus-locked negative deflections occurring at the mastoids peak at approximately the same latency as the occipital N160s, and thus somewhat earlier than those recorded from the central sites (Fig. 1). To conclude, we have shown that previously reported ipsilateral-contralateral asymmetries in the amplitudes and latencies of two VEP components elicited by small lateralized light flashes are unlikely to result either from light scatter or an inappropriate choice of reference site. To the extent that these data are considered informative about transcallosal processes (our reasons for considering them as such are elucidated in Lines et al. 1984 and Rugg et al. 1984a) this technique may provide a useful means for their assessment.
Summaff Visual evoked potentials (VEPs) to small lateralized flashes were recorded from the parietal midline, homotopic lateral central and occipital electrodes, and from left and right mastoid processes, all referred to a balanced non-cephalic reference. Two stimulus eccentricities, 4 ° and 10 °, were employed, and a comparison made between the non-cephalic and linked mastoid references. P120 (measured at lateral occipital sites only) peaked earlier and was of smaller amplitude at the electrode contralateral to the visual field of stimulus exposure. N160 peaked earlier at central than occipital sites, was larger from electrodes over the contralateral hemisphere, and at the occipital sites only, peaked earlier in the electrode contralateral
M,D. R U G G ET AL.
to the visual field of stimulus exposure. These effects were virtually unaffected by the eccentricity manipulation and it is concluded that light scatter across the visual midline is unlikely to be responsible for the observed pattern of ipsilateral-contralateral VEP asymmetries. The mastoids were found to detect significant stimulus-locked activity in the same latency range as the occipital N160 component. However, comparison of the non-cephalic and linked mastoid references revealed only non-specific effects, and no sign of any change in the pattern of ipsilateralcontralateral VEP asymmetries, or the magnitude of the associated latency differences. It is concluded that these effects may be of value in the study of callosal transfer.
R6sum6 Nouvelle &ude des potentiels bvoqubs v&uels par des stimulus latbralisks: effets de l'excentricitb du stimulus et du site de rbfbrence Des potentiels visuels 6voqurs (PEV) par des flash 16grrement latrralisrs furent enregistrrs par des 61ectrodes situres en parirtal mrdian, en centrale latrrale homotopique et occipitale, et sur les mastoides, toutes avec une rbfrrence non crphalique compensre. Deux excentricitrs du stimulus, de 4 ° et 10 °, furent utilisres et une comparaison entre les rrfrrences non c6phalique et mastoidienne rrunie a 6t6 faite. La P120 (mesurre aux sites occipitaux seulement) prrsentait un pic plus prrcoce et de plus faible amplitude h l'61ectrode contralatrrale au champ visuel expos6 au stimulus. N160 avait un pic plus prrcoce au sites centraux qu'occipitaux, 6tait plus grand pour les 61ectrodes sur l'hrmisph~re contralatrral au champ visuel soumis au stimulus. Ces effets n'rtaient virtuellement pas affectrs par la modification de l'excentricit6 et on en conclut qu'une dispersion de la lurnirre autour de la ligne visuelle mrdiane n'est vraisemblablement pas responsable du pattern observ6 d'asymrtrie des PEV ipsi- et contralatrraux. Les mastoides permettent de drtecter une activit6 significative li6e au stimulus dans le mrme
VEPs ELICITED BY LATERALIZED STIMULI d o m a i n e d e l a t e n c e q u e la c o m p o s a n t e o c c i p i t a l e N 1 6 0 . T o u t e f o i s , la c o m p a r a i s o n d e s r r f r r e n c e s n o n c r p h a l i q u e s et m a s t o ' i d i e n n e s r r u n i e s n ' a r r v r 1 6 que des effets non sprcifiques, et aucun signe de m o d i f i c a t i o n d a n s le p a t t e r n d ' a s y m r t r i e d e s P E V i p s i - et c o n t r a l a t r r a u x o u d a n s la g r a n d e u r d e s diffrrences de latences associres. On en conclut q u e ces e f f e t s p e u v e n t d t r e u t i l e s d a n s l ' r t u d e d e s transferts callosaux. This research is supported by the Wellcome Trust and the Medical Research Council of the U.K.
References Berlucchi, G. Recent advances in the analysis of the neural substrates of interhemispheric communication. In: O. Pompeiano and C. Ajmone Marsan (Eds.), Brain Mechanisms and Perceptual Awareness. Raven Press, New York, 1981: 133-152. Bruce, C.J., Desimone, R. and Gross, C.G. Visual properties of neurons in a polysensory area in superior temporal sulcus of the macaque. J. Neurophysiol., 1981, 46: 369-384. Gross, C.G., Bender, D.B. and Mishkin, M. Contributions of the corpus callosum and the anterior commissure to visual activation of inferior temporal neurons. Brain Res., 1977, 131: 227-239.
87 Halliday, A.M., Barrett, G., Halliday, E. and Michael, W.F. The topography of the pattern evoked potential. In: J.E. Desmedt (Ed.), Visual Evoked Potentials in Man: New Developments. Clarendon, Oxford, 1977: 121-133. Lepore, F. and Guillemot, J.-P. Visual receptive field properties of cells innervated through the corpus callosum in the cat. Exp. Brain Res., 1982, 46: 413-424. Lines, C.R., Rugg, M.D. and Milner, A.D. The effect of stimulus intensity on visual evoked potential estimates of interhemispheric transmission time. Exp. Brain Res., 1984, in press. Milner, A.D. and Lines, C.R. Interhemispheric pathways in simple reaction time to lateralized light flash. Neuropsychologia, 1982, 20: 171-179. Rugg, M.D. Electrophysiological studies. In: J.G. Beaumont (Ed.), Divided Visual Field Studies of Cerebral Organisation. Academic Press, London, 1982: 129-145. Rugg, M.D., Lines, C.R. and Milner, A.D. Visual evoked potentials to lateralised visual stimuli and the measurement of interhemispheric transmission time. Neuropsychologia, 1984a, 22: 215-225. Rugg, M.D., Milner, A.D. and Lines, C.R. Visual evoked potentials to lateralised stimuli in two cases of callosal agenesis. J. Neurol. Neurosurg. Psychiat., 1984b, in press. Stephenson, W.A. and Gibbs, F.A. A balanced non-cephalic reference electrode. Electroenceph. clin. Neurophysiol., 1951, 3: 237-240. Van Essen, D.C., Newsome, W.T. and Bixby, J.L. The pattern of interhemispheric connections and its relationship to extrastriate visual areas in the macaque monkey. J. Neurosci., 1982, 2: 265-283.