Vision Res. Vol. 34, No. 5, pp. 581-590, 1994 Copyright 0 1994Elsevier ScienceLtd Printed in Great Britain. All rights resewed 0042-6989/94 $6.00+ 0.00 Pergamon Hemispheric Asymmetry in the Maturation the Extrastriate Checkerboard Onset Evoked Potential P. OSSENBLOK,*T J. C. DE MUNCK,*$ H. J. WIERINGA,§ Received 22 June 1992; in revised form 21 December 1992; in final form of D. REITS,* H. SPEKREIJSE*l zyxwvutsrqponmlkjih 1 July 1993 Recently we have shown that the single positive deflection in the checkerboard onset evoked potential (EP) of young children of striate origin develops into a negative-positive complex. However, also an early positive peak becomes apparent in the checkerboard onset EP. To determine the origin and development of the activity underlying this early positive deflection we studied the checkerboard onset EPs in children of 9-16 years of age. It was found that for the children in this age group two different dipole sources are responsible for the activity underlying the pattern onset EP. One of the dipoles corresponds to the activity generated in the striate cortex, whereas a second dipole of extrastriate origin is responsible for the appearance of the early positive deflection. This extrastriate activity shows hemispheric asymmetry, i.e. the strength of the right hemispheric extrastriate source exceeds the strength of the left hemispheric source. These results are in accordance with histological studies of Cone1 (19394963) [The postnatal devefqment thehumun cerebral cortex (Vols l-g). Cambridge, Mass.: Harvard Univ. Press] which show that the maturation of the extrastriate areas of the left hemisphere is delayed with respect to the right hemisphere. of Checkerboard onset EP Equivalent dipole Maturation INTRODUCTION pattern onset evoked potential (EP) does not its characteristic positive-negative-positive obtain complex with adult-like CI, CII and CIII peaklatencies before puberty (Spekreijse, 1978; De VriesKhoe & Spekreijse, 1982; Spekreijse, 1983). In the first years of life the child checkerboard onset EP consists mainly of a single positive deflection (Spekreijse, 1978; De Vries-Khoe 8z Spekreijse, 1982), which has its origin in striate cortex (Ossenblok, Reits 8z Spekreijse, 1992). From the age of about 4 years on waveform changes of the pattern onset EP become apparent in the checkerboard onset EP, reflecting the growth of a negative peak (CII) which is preceded by an early positive peak (De Vries-Khoe & Spekreijse, 1982; Spekreijse, 1983; Apkarian, Reits & Spekreijse, 1984). It was shown by Ossenblok et al. The *The Netherlands Ophthalmic Research Institute, P.O. Box 12141, 1100 AC, Amsterdam-Zuidoost, The Netherlands. tPresent address: Department of Clinical Neurophysiology, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam-Zuidoost, The Netherlands. $The Netherlands Institute for Sea Research, Department of Physical Oceanography, P.O. Box 59, 1790 AB Den Burg, The Netherlands. §Faculty of Applied Physics, Low Temperature Department, University of Twente, P.O. Box 217,750O AE Enschede, The Netherlands. ~TO whom all correspondence should be addressed. Extrastriate activity Hemispheric asymmetry (1992) that the growth of the negative peak reflects changes in striate activity as a function of age. This paper deals with the origin and development of the activity underlying the early positive peak of the child checkerboard onset EP. The analysis is based on the assumption that the sources of the VEP can be modelled by equivalent dipole sources, which have a fixed position and orientation while the source strength is varying. Using the spatialtemporal dipole model of De Munck (1990) we studied the origin and development of the activity underlying the early positive peak of the pattern onset EP for children of 9-16 years of age. Since local contrast favours the early positive peak of the pattern onset EP (Spekreijse, Van der Tweel & Zuidema, 1973) we used relatively large checkerboard elements, high modulation depths and large stimulus fields to increase the signal-tonoise ratio of the responses. Histological studies of Cone1 (1939-1963) have shown that the maturation of the extrastriate areas of the left hemisphere is delayed with respect to the right hemisphere. We, therefore, compared also the results of left and right half-field stimulation systematically for all subjects studied. METHODS Stimuli Stimulation was by means of an electrostatic monochrome CRT (Hewlett Packard HP1321A) with a 581 582 P. OSSENBLOK resolution of 256*256 pixels and a frame rate of 107.76Hz. The CRT was driven by a digital display generator (Neuroscientific VENUS system). A/D acquisition rate was set to twice the frame rate (215.52 Hz) and was locked onto the frame rate of the CRT. The mean luminance of the CRT was 65 cd/m2. All stimuli were presented in the left or right hemifield surrounded by a steady homogenous field of the same mean luminance, with a field size subtending to 4 or 8 deg. The modulation depth of the stimulus is chosen such that the retinal contrast is constant at, respectively, a level of 25.3% or 52.1%. The retinal contrast was kept constant by means of the point spread function for the light scattering in the eye (Vos, 1984). The checkerboard pattern was presented for 300 msec every 800 msec, without net variation in overall luminance level. In this way the responses evoked by the onset and offset of the checkerboard could be registered separately. The results presented were obtained for both left and right half-field stimulation with check sizes of 24’. Recording VEPs were recorded at 24 scalp electrodes. The electrodes were fixed in a grid, with an inter-electrode distance of about 4 cm, covering the occipital region of the head and the actual positions were measured afterwards. A reference electrode was placed at the frontal midline; the ground was located near the vertex. Three reference points are required to define the reference frame of the electrodes. For this we used the two external auditory meati and the vertex. By positioning the legs of a pair of compasses at the electrode location and at the three reference points the location of every electrode Pi is uniquely defined. In this way the best fitting sphere for the electrode positions on the head is found (De Munck, Vijn & Spekreijse, 1991). The children were seated as comfortably as possible. The monitor was viewed binocularly from a distance of 86 cm. If necessary recording was interrupted by automatic artifact rejection or by an observer who monitored the child’s behaviour. Fixation was established with a set of three LEDs of different colour. This fixation point was presented in the centre of the screen and the subject had to respond verbally to each random appearance of a particular colour. To hold the attention of the children in between the recordings a children’s movie was presented. Signals were amplified (Medelec 5000) and bandpass filtered between 1.5 and 70 Hz. The high cut-off frequency (70 Hz) was set by a low-pass fourth order Butterworth filter, which introduces a phase shift increasing the response latencies by about 7 msec. Peaklatencies estimated from the recordings in this paper should be corrected for this latency increase. The EEGs were sampled with a 215.52 Hz sample frequency, and on-line averaged with the CED 1401 system. The signalto-noise ratio was estimated by means of the plus-minus averaging method (Schimmel, 1967). For the dipole analysis of the pattern onset EP a time window of 65.0-227.4 msec after pattern onset was chosen which PI </I includes the prominent responses. Equivalent maxima and minima of the dipole analysis Dipole sources give an adequate description of the scalp recorded potentials (De Munck, Van Dijk & Spekreijse, 1988a, b). To localize the dipoles the spatiotemporal dipole model of De Munck (1990) was used. In this model the problem of temporal overlapping source activity was solved by assuming a priori that the generators of the pattern onset EPs may be described by a fixed number of stationary dipoles, with time varying amplitudes. It is not known in advance how many dipoles are activated upon visual stimulation. Therefore a singular value decomposition (SVD) of the data is performed which provides us with an estimate of the minimal number of activated sources, by arguing that the number of significant principal components (PCs) equals the number of activated sources. The number of activated dipoles is chosen such that the lower limit of the residual, which depends solely on the number of significant PCs, reaches the noise level of the data. The starting locations of the dipoles used in the inverse calculation procedure are obtained by means of the moving dipole model. However, to be certain that the algorithm did not converge only to a local minimum different starting locations were tried, which all give the same solution. The dipoles are localized within the spherical three-shell model for the head, describing the brain surrounded by the skull and scalp (Ary, Klein & Fender, 1981), while no further constraints are placed on the volume searched. If the percentage of the recorded data variance not explained by the dipole solution reaches the lower limit of the residual the actual residual of the least square fit can be explained purely by noise. Note, however, that the actual residual obtained also may depend on model errors of the head and the sources. In this study the final solution of the inverse calculation is obtained if the difference between the actual residual obtained and the lower limit of the residual is minimal, which is reached if the partial derivative of the residual (see De Munck, 1990) is smaller as lo-‘. Subjects The results presented in this paper were obtained in 18 healthy children of 9-16 years of age recruited from employees of our institute. For all of these children the difference amplitude of the right and left hemifield responses is depicted. For comparison we also present the difference amplitude of nine adults, ranging from 21 to 47 years of age. For 13 of these children for whom the significant variance of the response sets evoked by both the right and left hemifield amounted to 80% of the total power of the responses or more also the results of the SVD are tabulated. Furthermore, for five of these children who are representative for the group, the results of the equivalent dipole analysis are depicted. Presentation of results The position and orientation of the dipoles are given with respect to the best fitting sphere (see Figs 2-4). MATURATION AK; (a) OF THE PATTERN ONSET EP 583 T(b) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 15.ayr AK ; 1 5 .8 yr 24’ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE 24’ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG FIGURE I. The visual evoked responses to the appearance of a 24’ checkerboard in the central 8 deg of the right half-field a) and of the left half-field (b). The responses are plotted according to a 2-dimensional projection of a spherical surface with iorizontal and vertical distances of approx. 4 cm. The asterisk indicates the position of the inion, which is located at the midline of the head, approx. 2cm above the bottom row of electrodes. Transformation over the x- and z-axis gives the in- head, above the inion, with an orientation which is ion-ear co-ordinate system. In Table 1 the parameters of partly radial and partly tangential. The location of the equivalent dipole indicates an origin in the striate cortex the dipoles are given with respect to this inion-ear system: the position parameters of the sources are given (area 17). Since the residual error of the dipole fit in the Cartesian co-ordinates X, Y and Z and the amounts to 2.5% of the total power of the responses the actual residual obtained can be explained largely orientation of the dipoles is given in the polar co-ordinates 8 and 4. The radial orientation is given as 8, by noise. A SVD of the responses evoked by left half-field whereas C$ yields the orientation in the tangential stimulation showed that the accumulated power of the plane. The residual error (R.err) equals the difference between the actual residual obtained and the lower limit two major principal components, amounting respectively to 82.0 and 14.4%, accounts for the significant of the residual. Note, furthermore, that in the computations of the source strength relative values for the radii power of the response set. Thus, two dipoles are needed and conductivities of the concentric shells describing the to account for the significant variance of the responses. volume conductor were used (Rush & Driscoll, 1968). zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG RESULTS The left part of Fig. 1 shows pattern EPs recorded at 24 scalp electrodes upon presentation of a pattern onset stimulus in the right half of the visual field. The responses are maximal at the midline of the head, above the inion. The wave form of the responses is characteristic for children of this age (Ossenblok et zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC al., 1992) and consists mainly of a negative-positive complex. For stimulation with a left half-field, however, also an early positive peak, at about 100 msec, becomes apparent in the responses [Fig. l(b)]. Note that this peak, which dominates at the contralateral recording site, is absent in the contralaterally recorded responses evoked by right half-field stimulation. The equivalent dipoles A SVD of the responses evoked by right half-field stimulation showed that the power of the first principal component (93.5%) accounts for the significant power of the responses within the time window chosen for the analysis. The variation in strength of the source is reflected by the component of the pattern onset EP [Fig. 2(a)]. Since the activity underlying this pattern onset EP can be described by a single source the shape of this component is, of course, similar to the time varying amplitude of the pattern onset EP. The corresponding dipole source is located near the midline of the J FIGURE 2. The locations and orientations of the sources are plotted in diagrams viewed from behind (left), from the right (middle) and from above (right). The location of the dipole source is referred to the best fitting sphere of the head whose outer limit is depicted. The length and direction of the arrow in each plane shows the strength and orientation of the dipole. In this figure the locations and orientations of the dipoles responsible for the responses shown in Fig. 1 are depicted. (a) The responses evoked by right half-field stimulation could be explained by a single source. (b) For left half-field stimulation two dipoles could be localized originating in the striate and extrastriate cortex. The shape of the components is shown for each dipole source. The components covered a time window that started at 65.0 msec after pattern onset and ended at 227.4msec. P. OSSENBLOK ef (I/. 5x4 (4 (b) = 24’- - 24’ I FIGURE 3. In this figure the parameters of the dipoles are depicted for three of the subjects studied, ranging from 11 to 16 years of age. The dipoles were activated by a 8 deg checkerboard with elements of 24’ for right and left half-field stimulation. The responses evoked by right half-field stimulation could be explained by a single source (a). For left half-field stimulation two dipoles could be localized (b) originating in the striate and extrastriate cortex. For further information on the presentation of the results see the legend of Fig. 2. The residual error of the dipole fit amounts to 1.0% of the total power of the response set, thus indicating that the actual residual obtained can be explained by noise. One of these dipoles is located near the midline of the head [Fig. 2(b)] and the strength profile of this source corresponds closely to the striate component found for stimulation of the right half of the visual field. The second dipole is located at the contralateral side of the head and away from the inion. This indicates an origin in one of the extrastriate areas of the visual cortex. The component corresponding to this dipole has a biphasic wave form with a positive deflection at 105msec. This component dominates the early positive peak of the pattern onset EP. The same picture emerges for three other subjects studied (Fig. 3). The significant variance of the responses generated by right half-field stimulation could always be accounted for by the activity of a single source, whereas TABLE 1. The parameters of the sources underlying the responses generated by right and left half-field stimulation with elements of 24’ for four of the subjects studied. The characters S and E indicate, respectively, a striate and extrastriate source. The Cartesian co-ordinates X, Y, and Z are given in cm and the polar coordinates 0 and 4 in deg. The last column gives the residual error of the dipole fit Right half-field stimulation X S/E MM MR AK SR Sl Sl Sl Sl 7.41 9.74 9.78 9.80 Left half-field stimulation x SIE MM MR AK SR Sl 1-g Sl IE Si IE Sl IE 7.58 6.76 9.36 8.49 8.90 7.14 7.40 6.94 Y Z 0 -0.55 0.33 -1.27 -2.48 3.60 2.19 3.38 4.50 60.6 54.8 56.6 68.1 Y Z 3.32 2.99 3.81 4.36 4.52 4.39 3.43 4.62 0.29 4.57 1.09 2.03 0.49 3.63 1.60 3.56 4 R.err 30.6 32.6 47.2 70.2 0.8% 2.1% 2.5% 2.3% 0 4 R.err 45.2 48.4 92.1 101.1 36.5 22.5 58.2 72.1 -6.8 -8.8 1.6 6.2 20.8 -24.8 -68.1 -61.7 1.3% 1.8% 1.O% 1.6% MATURATION OF THE PATTERN ONSET EP the responses evoked by left half-field stimulation needed two dipole sources. The striate sources generated by both right and left half-field stimulation are located near the midline of the head, about 2-5 cm ‘above the inion (Table 1). Note that the variation in strength of the striate sources generated by left half-field stimulation corresponds closely to the time profile of the striate response generated by right half-field stimulation. The extrastriate sources activated by left half-field stimulation are located contralaterally with a distance of about 2-5 cm from the midline of the head for all the subjects studied, whereas the orientation of the sources is partly radial and partly tangential (see Table 1). Their response profiles are quite similar for the three subjects studied with a peak-latency for the positive deflection around 100 msec followed by a negative deflection at about 160 msec, as shown before (Maier, Dagnelie, Spekreijse & Van Dijk, 1987; Ossenblok & Spekreijse, 1991). TABLE 2. The results of the SVD are listed for both the right(R) and left (L) hemifield for 13 of the children studied. The significant variance of the response sets (%SP) and the results of the SVD, obtained within the time window of analysis, are given as percentages of the total signal plus noise power. The power of the first and second principal component are denoted, respectively, as %PCI and %PC2 AGE FIELD 9.6 R L R L R L R L R L R L R L R L R L R L R L R L R T 10.6 11.0 11.1 11.2 11.8 12.6 12.7 Hemispheric asy mmetry 12.8 Thus we have shown that a single dipole is activated upon right half-field stimulation, whereas for left halffield stimulation two dipoles are needed to explain the significant variance of the responses. However, the hemispheric asymmetry is not always that clear. The results of the SVD show that for most of the children studied a second extrastriate source also may contribute to the responses evoked by right hemifield stimulation (Table 2). Note, however, that the power of the principal component corresponding to the right hemispheric extrastriate source (%PC2) exceeds these values of the corresponding left hemispheric source. For 1 ;ll.l yr FIGURE 4. The results described in the legends site 14 and the response numbers. Note that the 24’ 585 13.5 14.8 15.8 16.0 %SP %PCl %PC2 96.6 94.7 3.1 90.8 zyxwvutsrqponmlkjihgfedcba 80.3 16.4 96.6 89.9 7.2 93.7 67.7 25.7 90.0 87.3 7.7 92.1 69.7 21.1 90.5 65.6 26.3 91.5 55.3 40.1 93.7 90.6 4.3 97.2 82.7 13.7 84.8 82.9 13.6 88.9 26.0 66.1 94.3 82.6 12.1 88.3 72.8 24.6 94.5 80.1 13.5 93.1 76.2 16.6 96.2 71.2 27.4 96.5 31.3 65.7 98.1 64.4 34.2 92.5 52.7 40.2 95.6 92.2 4.8 94.7 16.3 80.5 95.1 3.9 93.5 96.4 14.4 82.0 94.2 91.9 4.6 o, 17.4 76.7 jr.7d instance, for the subject of 11.1 years old (KL), the results of the SVD show that two dipoles are needed to account for the significant variance of the responses zyxwvuts KL;1 1 .1 yr 24’ obtained for subject KL are presented. Stimulation and the presentation of results is the same as of Figs 1 and 2. The response evoked by right half-field stimulation and recorded at the electrode evoked by left half-field stimulation and recorded at electrode site 17 are marked with the respective mean strength of the extrastriate component generated by left half-field stimulation exceeds that of the extrastriate component generated by right half-field stimulation. P. OSSENBLOK et a/ . 586 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 30 a 16 20 age in years 30 50 FIGURE 5. (a) The difference amplitude is depicted for 18 subjects younger than 20 years of age and for 9 adult-subjects (~20 years) vs a logarithmic age scale: The vertical striped-dotted line separates the data points of both age groups. The difference potential was obtained by subtracting the left hemispheric response, recorded at electrode derivation 14, from the right hemispheric response recorded at electrode derivation 17 [see Fig. 4(a)]. Subsequently the amplitude of the diierence potential was obtained at the peak-latencies depicted in the bottom half of the figure. (b) The peak-late&es depicted vs the logarithmic age scale. These peak-latencies were measured at the incidence of the maximal amplitude of the first positive peak of the checkerboard onset EP. evoked by either half-field [Fig. 4(a)]. One of the dipoles has a striate origin, whereas a second dipole is located at the contralateral hemisphere [Fig. 4(b)]. Like for the other subjects studied the location and strength profile of the striate sources are quite alike for either half-field. The dipoles located contralaterally represent extrastriate sources. Note that in accordance with the results of the SVD the mean strength of the extrastriate right hemispheric source exceeds that of the left one. Since the activity of the extrast~ate source dominates the early positive peak of the pattern onset EP the hemispheric asymmetry can also be obtained by subtracting the left hemispheric response recorded at electrode 14 from the right hemispheric response recorded at electrode 17, since at these electrode sites the early positive peak of the responses is maximal [see Fig. 4(a)]. Figure 5(a) shows that the peak-amplitude of the d%%rence potential is positive for 17 of the 18 children studied, whereas for adults (> 20 years) the sign of the difference amplitude alternates around zero. A positive difference amplitude means that the right hemispheric response exceeds the left hemispheric response, while the difference amplitude in ~VS is an indication for the strength differences of the extrastriate sources. The difference amplitudes of the group of children studied and of the adult group differ signifrcantfy from each other (Mann-Whitney test, p < O.OOl), while the mean value of the difference amplitudes of children differs significantly from zero (p = 0.001). For children the mean value of the peak-later&es amounts to 98.1 2 2 msec, whereas the peak-latency of extrastriate activity found in adults amounts to a mean value of 94.2 4 1.1 msec [Fig. 5(b)]. So they do not diflFer significantly. The striate sources activated by left and rig& half-field stimulation are located both near the midfine of the head, above the inion (see Figs 2-4) and the time profiles AK; 15.8 yr FIGURE 6. The MRI-scans reconstruction also the inion part the left (L) hemisphere. section also the orientation, of one of the subjects studied (AK) were reconstructed such that a three-dimensional picture emerges of the head, with markers placed at the auditory meati. Since in this can be seen the best fitting sphere could be drawn with respect to the inion-ear system. The upper part of the horizontal section shows the right (R) hemisphere, the bottom The positions of the sources, indicated by white circles, are given with respect to this best fitting sphere for right (a) and left (b) half-field stimulation. In the horizontal indicated by the white line, is depicted. In the vertical section of the MRI-scan the position of the dipole and the intermting line in the horizontal plane is shown. Note that the best fitting sphere, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA shown in the horizontal section, fits the back of the head quite well. 588 P. OSSENBLOK et (I/. of the corresponding components look very much alike. whereas extrastriate activity starts to contribute later in The variation in strength of the extrastriate sources is life. While the maturation of striate activity is reflected in the growth of the negative peak (CII) of the pattern reflected in a positive-negative wave which dominates the early positive peak of the pattern onset EP. The onset EP, extrastriate activity dominates the early posiwaveforms of the extrastriate components of the differ- tive peak at about 100 msec. Thus not the ingrowth ent subjects studied look very much alike. Thus, wave- of the negative peak (CII) is responsible for the appearform changes as a function of age due to changes at the ance of this early positive peak, as was suggested by neuronal level, as reported for the striate component Spekreijse (1983) and Apkarian et al. (1984) but the contribution of area 18 activity to the pattern onset EP. (Ossenblok et al., 1992), are not apparent for the extrastriate component. Note that the strength profile of the Note, furthermore, that area 19 activity recorded in extrastriate source and its contralateral origin is identical adults (Ossenblok & Spekreijse, 1991) is absent in the onset EP. The regional developto the CI component of the adult response (Maier zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA et al., child checkerboard 1987) which has an origin in area 18 of the visual cortex ment of activity underlying the child checkerboard onset (Ossenblok & Spekreijse, 1991). Therefore the extrastriEP is in accordance with the regional development of ate source found in the child checkerboard onset EP has myelination patterns as shown by Cone1 (1939-1963) probably also an origin in area 18. Yakovlev and Lecours (1967) and Holland, Haas, Figure 3 shows, however, that the parameters of the Norman, Zawadzki and Newton (1986). The primary dipoles vary amongst the subjects studied. This may be visual cortex is the first to become myelinated at the due to source localization errors (Ary et al., 1981; Stok, age of about 7 years, therefore, probably, upon visual 1987; Van Dijk & Spekreijse, 1990) although a large part stimulation area 17 activity is the first to appear in the of the variability could be due to anatomical variation pattern onset EP (Ossenblok et al., 1992). The myelinaof the visual cortex (Brindley, 1972; Stensaas, Eddington tion of the secondary visual areas can continue to late & Dobelle, 1974; Steinmetz, Fiirst & Meyer, 1989). For puberty (area 18) or even beyond the second decade of one of the subjects studied (AK) the positions of the life (area 19) (Conel, 1939-1963; Yakovlev & Lecours, sources are depicted with respect to the anatomical 1967). Progress in myelination involves both the contristructures of the visual cortex, as shown in the MRIbution of regional activity to the VEP and a decrease in scans of Fig. 6. Note first that all three sources lay within peak-latency (McDonald, 1977; Halliday, McDonald & the grey matter of the cortex, representing brain tissue. Mushin, 1977). Note, however, that the peak-latencies of The horizontal section of the MRI-scans shows that the the extrastriate components found in children do not striate dipoles are located about 1 cm anteriorly and change significantly as a function of age, although the approximately mirror symmetric with respect to the variability is much larger than for the adult subjects medial plane for right and left half-field stimulation, [see Fig. 5(b)]. although the left hemispheric source is more eccentric. Thus the activity generated in area 18 of the visual Thus the position of these sources is in accordance with cortex still develops at the age of 16 years while area 19 the geometry of the primary visual cortex (Stensaas activity does not yet become apparent in the surface et al., 1974). The position of the striate sources in the potential. According to the delayed development of vertical direction also falls well within the anatomic area 18 activity in the child checkerboard onset EP variation of the striate cortex, since this part of the these subjects may show impaired perception of spatial cortex is located near the occiput, 34cm above the relations [see e.g. recent studies of Von der Heydt and inion, with an absolute variation in reference to the inion Peterhans (1989) and Peterhans and Von der Heydt of 4cm (Steinmetz et al., 1989). The distance from the (1989)], although children at the age of 16 years and medial plane of the source activated by left half-field younger show good visual performance. Note, however, that even the acuity of these children has not reached stimulation, which is indicated as an extrastriate source (see Fig. l), is about 4 cm as in adults (Ossenblok & adult levels yet (De Vries-Khoe & Spekreijse, 1982) thus Spekreijse, 1991). The MRI images show, furthermore, indicating that visual performance may still develop. that none of the sources are hidden within one of the Perception experiments in relation with electrophysiosulci of the visual cortex, thus accounting for the pre- logical results may provide a clue to answer the question to what visual deficits the absence of extrastriate activity dominant radial orientation of the dipoles. These results, therefore, provide further evidence (1) for the pre- in the pattern onset EP may lead. dominant radial orientation of the striate sources underlying the child checkerboard onset EP (see also Hemispheric asymmetry We have shown that extrastriate activity underlying Ossenblok et al., 1992) and (2) for an area 18 origin the child checkerboard onset EP shows hemispheric of the extrastriate sources underlying the child checkerboard onset EP, since the sources originating in area asymmetry. This asymmetry varies amongst subjects, probably because of the large variation in the state of 19 of the adult visual cortex are orientated tangentially development between the subjects in the age group (Ossenblok & Spekreijse, 1991). studied, although also the dominance of one of the Maturation of extrastriate activity hemispheres, as described by Stensaas et al. (1974) may affect the variance of the hemispheric asymmetry values. Striate activity dominates the checkerboard onset For some of the subjects a single dipole accounts for the EP of the youngest children (Ossenblok et al., 1992) MATURATION OF THE PATTERN ONSET EP 589 significant variance of the responses evoked by right (De Vries-Khoe 8c Spekreijse, 1982), while striate activity, which dominates the responses of children of half-field stimulation, whereas for other, even younger, subjects two dipoles are needed to describe the signifi- this age, is distributed symmetrically with respect to cant variance of the responses evoked by either right or the midline of the head (Ossenblok et al., 1992). zyxwvutsrqpon left half-field stimulation (see Figs 24). However, for many of the children studied something in between REFERENCES occurs: for these children two dipoles are needed to describe the significant variance of the left hemifield Apkarian, P. & Spekreijse, H. (1986). The VEP and misrouting responses, while a second although weak extrastriate pathways in human albinism. In Cracco, R. Q. & Bodis-Wollner, I. (Eds), Evoked potentials (pp. 211-226). New York: Alan R. Liss. source may also contribute to the right hemifield responses (see Table 2). Note, furthermore, that the Apkarian, P., Reits, D. & Spekreijse, H. (1983). A decisive electrophysiological test for human albinism. Electroencephalography and results of the SVD provide further evidence for Clinical Neurophy siology , 55, 5 13- 53 1. extrastriate hemispheric asymmetry. Apkarian, P., Reits, D. & Spekreijse, H. (1984). Component specificity A systematic comparison of the strength of the in albino VEP asymmetry: Maturation of the visual pathway anomaly. Experimental Brain Research, 53, 285- 294. extrastriate sources activated by right and left half-field stimulation is not always possible, because of Ary, J. P., Klein, S. A. & Fender, D. H. (1981). Location of sources of evoked scalp potentials: Correction for skull and scalp thickness. the constraints placed on the solution of the dipole IEEE Transactions on Biomedicinal Engineering, 28, 447452. source localization problem. Since, however, the early Brindley, G. S. (1972). The variability of the human striate cortex. positive peak of the pattern onset EP is dominated by Journal of Phy siology , 225, 1- 3P. extrastriate activity the difference amplitude of this Conel, J. L. (1939-1963). The postnatal development of the human cerebral cortex (Vols l- 8). Cambridge, Mass.: Harvard Univ. peak may provide an indication for the hemispheric Press. asymmetry of this activity. It was shown that the De Munck, J. C. (1990). The estimation of time varying dipoles on mean value of the difference amplitude recorded in the basis of evoked potentials. Electroencephalography and Clinical children from 9 to 16 years is significantly larger Neurophy siology , 77, 156160. than zero, thus reflecting extrastriate hemispheric asym- De Munck, J. C., Van Dijk, B. W. & Spekreijse, H. (1988a). Mathematical dipoles are adequate to describe realistic generators metry. This hemispheric asymmetry is in accordance of human brain activity. IEEE Transactions on Biomedicinal with histological studies of Cone1 (1939-1963) who Engineering, 35, 960- 966. showed that the axons of the pyramidal neurons in the De Munck, J. C., Van Dijk, B. W. & Spekreijse, H. (1988b). An extrastriate areas of the right hemisphere myelinate analytic method to determine the effect of source modelling errors on the apparent location and direction of biological sources. faster than in the left hemisphere. Since this hemispheric Journal of Applied Phy sics, 63, 944- 956. asymmetry can be estimated easily by subtracting the De Munck, J. C., Vijn, P. C. M. & Spekreijse, H. (1991). A maximal pattern onset EPs generated by, respectively, practical method for determining electrode positions on the head. right and left half-field stimulation, it may provide a Electroencephalography and Clinical Neurophy siology , 78, 8587. relative estimate of the maturation of the extrastriate De Vries-Khoe, L. H. & Spekreijse, H. (1982). Maturation of luminance and pattern EPs in man. Documenta Ophthalmologica checkerboard onset EP. For adults (>20 years) the Proceedings Series, 31, 461475. difference amplitude does not differ significantly from zero [see Fig. 5(a)]. Moreover, the alteration of the Ha&day, A. M., McDonald, W. I. & Mushin, J. (1977). Visual evoked potentials in patients with demyelinating disease. In Desmedt, sign of the strength difference around zero for the J. E. (Ed.), Visual woked potentials in man (pp. 438449). Oxford: adult subjects is in accordance with the arbitrarily Clarendon Press. distributed left or right hemispheric dominance as Holland, B. A., Haas, D. K., Norman, D., Zawadzki, M. B. & Newton, T. H. (1986). MRI of normal brain maturation, American shown by Stensaas et zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA al. (1974). Thus the hemispheric Journal of Neuroradiology , 7, 201- 208. asymmetry reflected in the pattern onset EP seems Maier, J., Dagnelie, G., Spekreijse, H. & Van Dijk, B. W. (1987). to disappear for subjects older than 20 years, which Principal components analysis for source localization of VEPs in should imply that from 20 years on the myolegenetic man. Vision Research, 27, 165- 177. cycle is complete for both hemispheres, at least up till McDonald, W. I. (1977). Visual evoked potentials in patients with demyelinating disease. In Desmedt, J.E. (Ed.), Visual evoked area 18. potentials in man (pp. 427438). Oxford: Clarendon Press. Apkarian, Reits and Spekreijse (1983) used the differOssenblok, P. & Spekreijse, H. (1991). The extrastriate generators ence amplitude of CI in diagnosing albinism. However, of the EP to checkerboard onset. A source localization approach. they never reported systematic CI magnitude differences Electroencephalography and Clinical Neurophy siology , 80, 181- 193. between both hemispheres. Note, however, that the Ossenblok, P., Reits, D. & Spekreijse, H. (1992). Analysis of striate activity underlying the pattern onset EP of children. Vision Research, asymmetry between the left and right eye topography, 32, 182991835. which is specific for albinos, remains apparent even when Peterhans, E. & Von der Heydt, R. (1989). Mechanisms of contour CI shows hemispheric amplitude differences. Thus the perception in monkey visual cortex. II. Contour bridging gaps. results of the study of Apkarian et al. (1983) are not Journal of Neuroscience, 9, 1749- l 763. contradictory to the present ones. Moreover, Apkarian Rush, S. & Driscoll, D. A. (1968). Current distribution in the brain from surface electrodes. Anesthesia and Analgesia Current and Spekreijse (1986) showed that the detection rate of Researches, 47, 727- 733. diagnosing albinism decreases sharply for children Schimmel, H. (1967). The (k) reference: Accuracy of estimated younger than 6 years of age. This is in accordance mean components in average response studies. Science, 157, 92- 93. with our results, since the early positive peak of the Spekreijse, H. (1978). Maturation of contrast EPs and development of pattern onset EP of young children is still immature visual resolution. Archives italiennes de biologic, 116, 358- 369. P. OSSENBLOK et al. 590 Spekreijse, H. (1983). Comparison of acuity tests and pattern evoked Von der Heydt, R. & Peterhans, E. (1989). Mechanisms of contour potential criteria: Two mechanisms underlay acuity maturation in perception in monkey visual cortex. I. Lines of pattern discontinuity. man. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Behavioral Brain Research, IO, 107-I 17. Journal of Neuroscience, 9, 1731l1748. Spekreijse, H., Van der Tweel, L. H. & Zuidema, T. (1973). Contrast Vos, J. J. (1984). Disability glare-A state of the art report. evoked responses in man. Vision Research, 13, 157771601. CIE-journal, 3. 39953. Steinmetz, H., Fiirst, G. & Meyer, B. (1989). Craniocerebral topograYakovlev, P. I. & Lecours, A. (1967). The myelogenetic cycles of phy within the international IO-20 system. Electroencephalography regional maturation of the brain. In Minkowski A. (Ed.), and Clinical Neurophy siology , 72, 499- 506. Regional development of the brain in early hfe (pp. 3 70). PhiladelStensaas, S. S., Eddington, D. K. & Dobelle, W. H. (1974). The phia: Davis. topography and variability of the primary visual cortex in man. Journal of Neurosurgery , 40, 7477755. Stok, C. J. (1987). The influence of model parameters on EEG/MEG single dipole source estimation. IEEE Transactions on Biomedical Engineering, 34, 289- 296. Van Dijk, B. W. 8c Spekreijse, H. (1990). Localization of electric and magnetic sources of brain activity. In Desmedt, J. E. (Ed.), Visual evoked potentials (pp. 57-74). Amsterdam: Elsevier. Acknowledgements- The software for the acquisition and presentation of the data was partly developed by Peter Vijn and Victor Lamme. The authors gratefully acknowledge, furthermore, the assistance of Marjorie Gilhuijs who as a student temporarily joined this research project.