- Email: [email protected]
Morphology of transient VEPs to luminance and chromatic pattern onset and offset
Vision Research 39 (1999) 1577 – 1584
Morphology of transient VEPs to luminance and chromatic pattern onset and offset Catherine M. Suttle a,*, Graham F.A. Harding b a
b
School of Optometry, Uni6ersity of California, Berkeley, CA 94720 -2020, USA Clinical Neurophysiology Unit, Aston Uni6ersity, Aston Triangle, Birmingham, B4 7ET, UK Received 29 January 1998; received in revised form 4 June 1998
Abstract Characteristics of the visual evoked response to chromatic and luminance-modulated stimuli reflect the activity of underlying neural mechanisms, although selective neuronal activity depends upon stimulus parameters. In the present study, the behaviour of the transient visual evoked response to low spatial and temporal frequency chromatic stimuli is investigated at a range of colour luminance ratios. Our results show that the response to pattern-offset may be used in addition to the pattern-onset response as part of the signature of the evoked response to luminance-modulated or isoluminant chromatic stimuli. © 1999 Elsevier Science Ltd. All rights reserved. Keywords: Chromatic; Luminance; VEP; Morphology
1. Introduction Transient visual evoked potentials (VEPs) to patternonset luminance and chromatic stimuli have been found to differ in morphology, being dominated by a positive peak component in response to luminance and a negativity in response to chromatic stimulation (Murray, Parry, Carden & Kulikowski, 1987; Berninger, Arden Hogg & Frumkes, 1989; Kulikowski, Murray & Parry, 1989; Rabin, Switkes, Crognale, Schneck & Adams, 1994). Low spatial frequency stimuli presented by pattern-reversal, however, elicit responses to chromatic and luminance-modulation of similar polarity, being positive-going peak response components (Regan & Spekreijse, 1974; Kulikowski et al., 1989), i.e. for achromatic stimuli, onset and reversal presentation elicit VEPs dominated by components of similar polarity; for low spatial frequency chromatic stimulation, pattern-onset and reversal VEPs are of opposite polarity. Low spatial frequency achromatic stimuli elicit responses from cortical cells stimulated by contrast * Corresponding author. Tel.: +1 510 6427679; fax: + 1 510 6435109; e-mail: [email protected].
change, while cells responsive to chromatic stimuli and high spatial frequency achromatic stimuli detect absolute contrast (Kulikowski, 1991). Thus, colour-opponent and broad-band cells both respond well to the appearance of a pattern, while broad-band (luminance) but not chromatic cells respond well when contrast of the pattern is eliminated at pattern-offset (Berninger et al., 1989). Components of magnocellular and parvocellular origin have been elicited in evoked responses to luminance-modulated and chromatic stimuli, when appropriate stimulus parameters are applied (McKeefry, Russell, Murray & Kulikowski 1996; Baseler & Sutter, 1997). Cells of the magnocellular (M) pathway prefer luminance-modulated stimuli of low spatial and high temporal frequency and are important for motion perception, while those of the parvocellular (P) pathway are more sensitive to chromatic contrast, to luminance-modulated stimuli of high spatial frequency and to low temporal frequency stimuli (Schiller et al., 1990; Anderson, Holliday, Singh & Harding, 1996). Cells of both M and P pathways respond to achromatic stimuli of moderate to high contrast (above 0.1) while at low contrasts, luminance-modulated and chromatic grating stimuli elicit responses dominated by the activity of M and P cells, respectively (Kulikowski et al., 1989; Kulikowski, Murray & Russell, 1991).
0042-6989/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved. PII: S0042-6989(98)00223-5
1578
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
Magnocellular and parvocellular pathways show considerable functional overlap, however (Logothetis et al., 1990; Schiller, 1991) and complete isolation is never assured, but may be maximized by using carefully chosen spatial and temporal stimulus parameters (Murray & Kulikowski, 1983; Kulikowski et al., 1989; Kulikowski et al., 1991; Kulikowski, Robson & McKeefry, 1996). Low to medium spatial frequencies allow both systems to respond to chromatic and achromatic stimuli, while higher spatial frequencies bias responses to achromatic stimuli towards the parvocellular channel, and very low spatial frequencies bias the response towards the magnocellular channel (Kulikowski et al. 1989). Moderate to high spatial frequency stimuli with a large number of spatial cycles (more than eight to ten for red-green stimuli) may introduce chromatic aberration, with the result that only one of the two spatial sinusoids is brought into sharp focus (Flitcroft, 1989; Kulikowski, 1991). In addition, large stimulus fields may introduce variations in isoluminance with eccentricity (Zrenner, 1983). When low contrast stimuli are used, however, responses to large or small fields show similar characteristics (Kulikowski et al., 1991) suggesting that contrast control is more important than field size for the selective stimulation of magnocellular and parvocellular channels of the visual system. It has been noted that chromatic cells respond poorly to chromatic stimulus offset (Jankov, 1978; Zrenner, 1983). The pattern-offset response to chromatic stimuli appears to be present when moderate contrast (0.44) is used, and diminished under lower (0.22) contrast conditions (Kulikowski et al., 1991) (see their Fig. 2). For achromatic stimuli, the response to pattern-offset diminishes with increasing spatial frequency, suggesting that the offset response is weakened when elicited from
P-cells, responsive to high spatial frequency patterned stimuli (Murray & Kulikowski, 1983). In the present study, transient visual evoked responses to high contrast luminance-modulated and chromatic stimuli of low spatial and temporal frequency were recorded from four human adults, all aged under 35 years: three colour-normal females and one protanomalous male. The principal aim of the study was to observe characteristics of the evoked response to large field, high contrast chromatic stimulation as luminance differences are introduced (high contrast and large field stimuli were used with the intention of comparison with responses to similar stimuli recorded in infants, not reported here) (Suttle, Anderson & Harding, 1997). Our results demonstrate a lack of response to pattern-offset of isoluminant chromatic stimuli, a strong response to pattern-offset of luminance-modulated stimuli, and distinct morphology of the response to onset of luminance and chromatic patterns. These findings, with non-optimal stimuli, appear to reflect the characteristics of underlying physiological mechanisms, with luminance-responsive cells detecting contrast change, and chromatic-sensitive cells responding to the absolute level of contrast. 2. Methods
2.1. Subjects Three female (AE, FF and VT) and one male (IP) adult were included in the study, age range 26–34 to avoid age-related changes in ocular media density (Boettner & Wolter, 1962). For each subject, colour vision was tested using the Ishihara pseudoisochromatic plates and the Farnsworth–Munsell 100-Hue test under a standard illuminant C. The 100-Hue and Ishihara test indicated protanomaly for this subject (Farnsworth, 1943; Kinnear, 1970). The three female subjects all demonstrated normal colour vision on both tests.
2.2. Stimuli
Fig. 1. Colour luminance ratios representing individual isoluminant points for three colour-normal observers (FF, AE and VT) and one protanomalous observer (IP; red-green match only carried out in this case), determined by heterochromatic flicker photometry at a temporal frequency of 20 Hz. Photometric isoluminance is at a ratio of 0.5.
, tritan matches; , red-green matches.
Horizontal, luminance-modulated and chromatic (green–purple or red–green) sinusoidal gratings of spatial frequency 1.0 cpd were generated using a VSG 2/2 graphics board (Cambridge Research Systems) with 14-bit DACs. Low spatial frequency was chosen to avoid chromatic aberrations. Gratings were displayed on a 17-in. diameter CRT colour monitor (Eizo flexscan 562-T) at a non-interlaced frame rate of 150 Hz and line scan rate of 66 kHz. The resolution of the screen was 380 lines× 800 pixels. The monitor was viewed binocularly at a distance of 1 m, which allowed 14 cycles of sinusoid to be displayed. The angular subtense of the display was 18.5° horizontally×14° vertically.
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
1579
Fig. 2. Visual evoked responses to red-green (A) and tritan (B) stimuli at the full range of colour luminance ratios recorded from colour-normal adult VT. The points of pattern onset and offset are indicated by arrows on the abscissa.
The red, green and blue colour guns of the monitor were gamma-corrected, to ensure that luminance output was proportional to voltage input, using Cambridge Research Systems’ VSG software. Calibration studies showed that the phosphor chromaticity co-ordinates for each gun remain stable for at least 2 h, beginning 10 min after switching the monitor on. Recordings were always completed well within this time window. Chromatic stimuli were presented along two axes in CIE colour space: a tritanopic confusion axis intended to stimulate selectively the S-cone pathway, and an axis orthogonal to this to stimulate the L – M pathway. Both axes passed through a white point at x = y =0.31. Chromaticity co-ordinates were x= 0.36, y = 0.45 (greenish–yellow) and x =0.27, y= 0.22 (purple) for the tritan stimulus, and x =0.38, y =0.275 (red) and
x= 0.23, y= 0.35 (green) for the red-green stimulus. Chromaticity co-ordinates were measured using the Bentham PMC 3B spectroradiometer (Bentham Instruments, UK). Isoluminance of each colour scheme was determined in each of the three colour-normal adults, using heterochromatic flicker photometry. Red–green or green– purple stimuli were phase-reversed at a temporal frequency of 20 Hz. Isoluminance was similar for each colour-normal adult, and closely approximated photometric isoluminance (see Fig. 1). For this reason, stimuli were presented to each adult at a range of colour luminance ratios centered about photometric isoluminance (ratio 0.5). Colour luminance ratio values used here indicate the ratio of green luminance to total (green+ purple) luminance in the tritan stimulus, and
1580
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
Fig. 2. (Continued)
red luminance to total (red +green) luminance in the red – green stimulus. The ratios used were 0.0, 0.2, 0.4, 0.45, 0.5, 0.55, 0.6, 0.8 and 1.0. Thus, over the full range of colour luminance ratios, the red-green stimulus, e.g. would change from green – black through red– green to red–black. The mean luminance of the display was maintained at 20 cpd/m2, for all stimuli, and was measured using a Minolta LS 110 photometer. For the tritan stimulus at photometric isoluminance, cone contrast was 0.0, 0.0 and 0.82 for L- M- and S-cones, respectively. For the isoluminant red-green stimulus, these values were 0.15, 0.28 and 0.04, respectively (Cole & Hine, 1992). These stimuli were at the maximum contrast level available using our monitor; VEP threshold is at approximately 6 and 9% of these levels (Rabin et al., 1994). Michelson contrast of the luminance-modulated stimuli (colour ratio 0.0 or 1.0) was 0.9.
Abrupt pattern-onset presentation was used at a temporal frequency of 2 Hz, with a stimulus duration of 100 ms and inter stimulus interval of 400 ms, during which the stimulus was replaced by a blank screen of the same mean chromaticity and luminance. This temporal duty cycle was chosen to maximize the response amplitude to onset-offset stimuli (Rabin et al., 1994). Onset-offset presentation was chosen in preference to pattern-reversal to allow a clear morphological distinction between responses to chromatic and luminance stimuli (Murray et al., 1987; Berninger et al., 1989; Kulikowski et al., 1989; Rabin et al., 1994). Stimuli were presented at the full range of colour ratios in pseudo-random order. Responses were recorded to both types of chromatic stimuli from the three colournormal subjects, and to only the red-green stimuli from the colour-defective subject (IP).
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
2.3. VEP recording Transient visual evoked responses were recorded from Oz of the International 10 – 20 system, referred to Cz and Fz was used as an earth. Signals were bandpass filtered (1–50 Hz) and 50 sweeps were averaged using the Medelec Sapphire 4E. Silver-silver chloride 1 cm diameter electrodes were used. Omniprep skin preparation was used on each electrode position before fixing with Blenderm tape. Resistances were always below 8 kV.
3. Results Fig. 2 shows visual evoked responses to tritan and red-green stimuli at the full range of colour luminance ratios for colour-normal subject VT. The responses shown here are typical of responses recorded from all three colour-normal subjects. At photometric isoluminance (colour ratio 0.5) the peak response component is a negativity occurring approximately 150 ms or 120 ms after stimulus onset for tritan and red-green, respectively, followed by a smaller positivity (around 100 ms later, at 220–240 ms after stimulus onset) in agreement with previous findings (Berninger et al., 1989; Rabin et al., 1994). As luminance differences are introduced to the chromatic stimulus, a second negative component occurs on the descending limb of the onset response, at a latency of approximately 170 ms for both types of chromatic stimuli. This second negativity is not apparent at isoluminance, and becomes better defined with luminance difference. Concurrent with this, the earlier negativity, which is the peak response component at isoluminance, becomes diminished as luminance differences are introduced to the stimulus. At colour ratios of 0.0 and 1.0 (luminance-modulated stimuli) this earlier negative component is only vaguely apparent, and is slightly later than the peak response to luminance stimulus onset, a small positivity at 80 – 100 ms latency. The clearest component in the luminance response (colour ratios 0.0 and 1.0) is the response to stimulus offset, a P-N-P component, the negativity of which occurs approximately 130 ms after offset. Fig. 3 shows the evoked response to the luminance-modulated stimulus (colour ratio 0.0) with offset having occurred at 100 ms, as for all other responses here, and in the lower trace with offset at 200 ms. The P-N-P component occurs correspondingly later, confirming that this is a response to stimulus offset. The response to stimulus offset also demonstrates changing morphology as luminance ratio of the chromatic stimulus is varied. At isoluminance, the response to red-green stimuli shows a very small P-N-P component, the negativity of which occurs approximately 230 ms after stimulus offset. As luminance differences are
1581
introduced, this component increases in amplitude and definition until at colour ratios of 0.0 and 1.0 the response to stimulus offset is the clearest component in the evoked response. The tritan response also shows this morphological change, with a response to offset becoming clearer at colour ratios away from isoluminance, although the trend is less clear than in the red-green responses. As demonstrated by Fig. 2, the evoked responses to stimulus onset and offset show opposing morphological changes towards isoluminance. The onset response becomes better defined, and the offset response diminishes, as isoluminance is approached. Fig. 4 illustrates this trend in terms of onset and offset response amplitudes, at the full range of colour luminance ratios, for all three colour-normal observers. Evoked responses recorded from protanomalous subject IP to red-green stimuli are shown in Fig. 5. At a colour luminance ratio of 0.6 the evoked response shows no clear components. This ratio is close to red–green isoluminance for this subject, as determined by heterochromatic flicker photometry (see Fig. 1). At all other ratios, a small positivity is seen at approximately 130 ms after stimulus onset. In addition, the P-N-P response to stimulus offset is apparent 120–130 ms after stimulus offset for responses to colour ratios other than 0.6.
4. Discussion When onset–offset presentation is used, peak responses to luminance and chromatic pattern-onset differ in polarity indicating that responses to luminance and chromatic pattern onset are dominated by, but not
Fig. 3. The visual evoked response to stimulus offset is demonstrated here by doubling the onset duration to 200 ms, so that stimulus offset occurs 100 ms later in the lower than in the upper trace (colour-normal adult FF). Onset and offset times are indicated on the abscissa.
1582
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
Fig. 4. Amplitude of the evoked potential to stimulus onset (A) and offset (B) is shown for all three colour-normal subjects, at the full range of colour luminance ratios. In (B), points at zero indicate that no component could be distinguished from the response. Peak to peak amplitudes were measured from the peak of the negative component to the following positivity of the onset response, and from positive peak to negativity of the offset response. The figure legend relates symbols to subjects, identified by their initials.
elicited exclusively from, distinct groups of cortical cells (see Section 1). In the present study, the peak response to chromatic onset stimulation was found to be negative-going, and the peak response to luminance onset was a positivity, in agreement with previous findings (Murray et al., 1987). While these responses are unlikely to be entirely selective, response morphology suggests that the responses to chromatic stimulation at
photometric isoluminance are dominated by the activity of colour-selective cortical cells, and that responses to luminance modulation are dominated by broad-band (luminance-responsive) cells. As luminance differences were introduced to the chromatic stimulus, a second negativity appeared in the peak response to chromatic pattern-onset, being absent or minimal at isoluminance and becoming increasingly
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
well-defined at colour ratios away from isoluminance (see Fig. 2). This morphological change suggests that the second component, elicited as luminance differences were introduced to the stimulus, is a response not to chromatic contrast but to luminance modulation. A similar change in morphology has been observed previously in evoked responses to chromatic stimuli (Kulikowski et al., 1989). In that study, the peak response to chromatic pattern onset was found to consist of two negative components, the earlier of which was found to be related to chromatic contrast, and the second to pattern elements of the stimulus, because this component increased in amplitude as the stimulus changed from one- to two-dimensions, and because a similar component was present in the achromatic response. The evoked response to stimulus offset is minimal at photometric isoluminance and becomes increasingly well-defined as luminance differences are introduced to the stimulus. Chromatic cells respond weakly to stimulus offset (Jankov, 1978; Zrenner, 1983), responding to absolute contrast rather than to contrast change. The contrast of the stimulus at offset is zero, and hence the response to this stimulus from chromatic cells is ex-
Fig. 5. Visual evoked responses to red-green stimuli at the full range of colour luminance ratios are shown, recorded from protanomalous adult IP. Pattern onset and offset times are indicated on the abscissa. The red-green isoluminant point for this observer determined by heterochromatic flicker photometry was 0.55.
1583
pected to be zero. Luminance-responsive cells respond to contrast change, and so these cells respond to both onset and offset phases of the stimulus. These response patterns are reflected in the evoked responses recorded in the present study, with the response to isoluminant chromatic stimulation showing minimal offset response, and the response to isochromatic luminance stimulation showing maximal offset response. At colour luminance ratios either side of photometric isoluminance, the offset response begins to appear, and becomes more well defined with increasing luminance difference. The small response to stimulus offset seen in the evoked response to isoluminant red–green stimulation (Fig. 2) may reflect the fact that the stimulus was at photometric isoluminance, a slight deviation from the individual’s isoluminant point (see Fig. 1). Visual evoked responses have been widely used to investigate colour vision deficiencies. In general, a stimulus along an appropriate colour confusion axis elicits diminished or absent evoked responses from a dichromatic or anomalous trichromatic observer (Kinney & McKay, 1974; Regan & Spekreijse, 1974; Crognale, Switkes, Rabin & Schneck, 1993). In the present study, the evoked response of a protanomalous observer to red–green stimulus pattern-onset is unclear at all colour luminance ratios, and the response to patternoffset is absent at a colour luminance ratio of 0.6 (see Fig. 5), close to this observer’s isoluminant point (Fig. 1). At ratios either side of this, the evoked response to stimulus offset appears and becomes more defined at ratios away from this point, presumably reflecting the increasing activity of luminance-responsive cells. At a R/R + G colour luminance ratio of 0.6, the red grating would appear brighter than the green to a normal trichromatic observer, but is close to perceived isoluminance for a protanomalous observer (Allen, Banks & Norcia, 1993). At this point the red and green components of such a stimulus, designed to modulate L- and M-cone types, provide equal stimulation of the remaining functional cone type being stimulated. Using the transient visual evoked response to low spatial and temporal frequency tritan and red-green chromatic stimuli at a range of colour luminance ratios spanning photometric isoluminance, we have demonstrated morphological changes to the onset and offset response components as luminance differences are introduced to the chromatic stimulus. Such changes to the onset response with colour luminance ratio are well known, and have been used to demonstrate the involvement of magnocellular and parvocellular pathways in the evoked response (Murray et al., 1987). It is also well understood that chromatic cells respond poorly to stimulus offset (Jankov, 1978; Zrenner, 1983). The results of the present study show that at photometric isoluminance the response to stimulus offset is minimal or absent, and that at colour luminance ratios removed
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
1584
from isoluminance this response component begins to appear, and becomes more apparent and well defined with increasing luminance difference. In one protanomalous observer, the response to chromatic contrast was found to be minimal, and the response to pattern-offset showed increasing definition with luminance difference, as also seen in the responses from colour-normal observers. These results demonstrate that morphological changes to the response component to pattern-offset may be used in addition to pattern-onset response characteristics, in colour-normal and colour-defective observers, as part of the signature of the transient evoked response to chromatic stimuli at isoluminance.
Acknowledgements CMS thanks the ISCEV Board of Directors for a grant to travel to the 1996 annual symposium, where this work was originally presented.
References Allen, D., Banks, M. S., & Norcia, A. M. (1993). Does chromatic sensitivity develop more slowly than luminance sensitivity? Vision Research, 33, 2553 – 2562. Anderson, S. J., Holliday, I. E., Singh, K. D., & Harding, G. F. (1996). Localization and functional analysis of human cortical area V5 using magneto-encephalography. Proceedings of the Royal Society of London. Series B: Biological Sciences, 263, 423 – 431. Baseler, H. A., & Sutter, E. E. (1997). M and P components of the VEP and their visual field distribution. Vision Research, 37, 675 – 690. Berninger, T. A., Arden, G. B., Hogg, C. R., & Frumkes, T. (1989). Separable evoked retinal and cortical potentials from each major visual pathway: preliminary results. British Journal of Ophthalmology, 73, 502 – 511. Boettner, E. A., & Wolter, J. R. (1962). Transmission of the ocular media. In6estigati6e Ophthalmology, 1, 776–783. Cole, G. R., & Hine, T. (1992). Computation of cone contrasts for color vision research, behavior research methods. Instruments and Computers, 24, 22 – 27. Crognale, M., Switkes, E., Rabin, J., & Schneck, M. E. (1993). Application of the spatio-chromatic visual evoked potential to detection of congenital and acquired colour deficiencies. Journal of the Optical Society of America A, 10, 1818–1825. Farnsworth, D. (1943). The Farnsworth-Munsell 100-Hue and dichotomous tests for color vision. Journal of the Optical Society of America, 33, 568 – 578.
.
Flitcroft, D. I. (1989). The interactions between chromatic aberration, defocus and stimulus chromaticity: implications for visual physiology and colorimetry. Vision Research, 29, 349 – 360. Jankov, E. (1978). Spectral sensitivity of off-response in human VECP during selective chromatic adaptation (author’s translation). Albrecht 6on Graefes Archi6 fur Klinische und Experimentelle Ophthalmologie, 206, 121 – 133. Kinnear, P. R. (1970). Proposals for scoring and assessing the 100Hue test. Vision Research, 10, 423 – 433. Kinney, J. A., & McKay, C. L. (1974). Test of color-defective vision using the visual evoked response. Journal of the Optical Society of America A, 64, 1244 – 1250. Kulikowski, J. J., Murray, I. J., & Parry, N. R. A. (1989). Electrophysiological correlates of chromatic-opponent and achromatic stimulation in man. In B. Drum, & E. Verriest, Colour 6ision deficiencies IX. Dordrecht: Academic, 145 – 153. Kulikowski, J. J., Murray, I. J., & Russell, M. H. A. (1991). Effect of stimulus size on chromatic and achromatic VEPs. In B. Drum, J. D. Moreland, & A. Serra, Colour 6ision deficiencies X. Dordrecht: Academic, 51 – 56. Kulikowski, J. J. (1991). On the nature of visual evoked potentials, unit responses and psychophysics. In A. Valberg, & B. B. Lee, From pigments to perception. New York: Plenum, 197 – 208. Kulikowski, J. J., Robson, A. G., & McKeefry, D. J. (1996). Specificity and selectivity of chromatic visual evoked potentials. Vision Research, 36, 3397 – 3401. Logothetis, N. K., Schiller, P. H., Charles, E. R., & Hurlbert, A. C. (1990). Perceptual deficits and the activity of the color-opponent and broad-band pathways at isoluminance. Science, 247, 214– 217. McKeefry, D. J., Russell, M. H., Murray, I. J., & Kulikowski, J. J. (1996). Amplitude and phase variations of harmonic components in human achromatic and chromatic visual evoked potentials. Visual Neuroscience, 13 (4), 639 – 653. Murray, I. J., & Kulikowski, J. J. (1983). VEPs and contrast. Vision Research, 23, 1741 – 1743. Murray, I. J., Parry, N. R. A., Carden, D., & Kulikowski, J. J. (1987). Human visual evoked potentials to chromatic and achromatic gratings. Clinical Vision Sciences, 1, 231 – 244. Rabin, J., Switkes, E., Crognale, M., Schneck, M. E., & Adams, A. J. (1994). Visual evoked potentials in three-dimensional color space: correlates of spatio-chromatic processing. Vision Research, 34, 2657 – 2671. Regan, D., & Spekreijse, H. (1974). Evoked potential indications of colour blindness. Vision Research, 14, 89 – 96. Schiller, P. H., Logothetis, N. K., & Charles, E. R. (1990). Functions of the color-opponent and broad-band channels of the visual system. Nature, 343, 68 – 70. Schiller, P. H. (1991). The color-opponent and broad-band channels of the primate visual system. In A. Valberg, & B. B. Lee, From pigments to perception. New York: Plenum Press. Suttle, C. M., Anderson, S. J., & Harding, G. F. A. (1997). A longitudinal study of visual evoked responses to tritan stimuli in human infants. Optometry and Vision Science, 74 (9), 717–725. Zrenner, E. (1983). Neurophysiological aspects of color 6ision in primates. Berlin: Springer-Verlag Chapter 6.
Morphology of transient VEPs to luminance and chromatic pattern onset and offset Catherine M. Suttle a,*, Graham F.A. Harding b a
b
School of Optometry, Uni6ersity of California, Berkeley, CA 94720 -2020, USA Clinical Neurophysiology Unit, Aston Uni6ersity, Aston Triangle, Birmingham, B4 7ET, UK Received 29 January 1998; received in revised form 4 June 1998
Abstract Characteristics of the visual evoked response to chromatic and luminance-modulated stimuli reflect the activity of underlying neural mechanisms, although selective neuronal activity depends upon stimulus parameters. In the present study, the behaviour of the transient visual evoked response to low spatial and temporal frequency chromatic stimuli is investigated at a range of colour luminance ratios. Our results show that the response to pattern-offset may be used in addition to the pattern-onset response as part of the signature of the evoked response to luminance-modulated or isoluminant chromatic stimuli. © 1999 Elsevier Science Ltd. All rights reserved. Keywords: Chromatic; Luminance; VEP; Morphology
1. Introduction Transient visual evoked potentials (VEPs) to patternonset luminance and chromatic stimuli have been found to differ in morphology, being dominated by a positive peak component in response to luminance and a negativity in response to chromatic stimulation (Murray, Parry, Carden & Kulikowski, 1987; Berninger, Arden Hogg & Frumkes, 1989; Kulikowski, Murray & Parry, 1989; Rabin, Switkes, Crognale, Schneck & Adams, 1994). Low spatial frequency stimuli presented by pattern-reversal, however, elicit responses to chromatic and luminance-modulation of similar polarity, being positive-going peak response components (Regan & Spekreijse, 1974; Kulikowski et al., 1989), i.e. for achromatic stimuli, onset and reversal presentation elicit VEPs dominated by components of similar polarity; for low spatial frequency chromatic stimulation, pattern-onset and reversal VEPs are of opposite polarity. Low spatial frequency achromatic stimuli elicit responses from cortical cells stimulated by contrast * Corresponding author. Tel.: +1 510 6427679; fax: + 1 510 6435109; e-mail: [email protected].
change, while cells responsive to chromatic stimuli and high spatial frequency achromatic stimuli detect absolute contrast (Kulikowski, 1991). Thus, colour-opponent and broad-band cells both respond well to the appearance of a pattern, while broad-band (luminance) but not chromatic cells respond well when contrast of the pattern is eliminated at pattern-offset (Berninger et al., 1989). Components of magnocellular and parvocellular origin have been elicited in evoked responses to luminance-modulated and chromatic stimuli, when appropriate stimulus parameters are applied (McKeefry, Russell, Murray & Kulikowski 1996; Baseler & Sutter, 1997). Cells of the magnocellular (M) pathway prefer luminance-modulated stimuli of low spatial and high temporal frequency and are important for motion perception, while those of the parvocellular (P) pathway are more sensitive to chromatic contrast, to luminance-modulated stimuli of high spatial frequency and to low temporal frequency stimuli (Schiller et al., 1990; Anderson, Holliday, Singh & Harding, 1996). Cells of both M and P pathways respond to achromatic stimuli of moderate to high contrast (above 0.1) while at low contrasts, luminance-modulated and chromatic grating stimuli elicit responses dominated by the activity of M and P cells, respectively (Kulikowski et al., 1989; Kulikowski, Murray & Russell, 1991).
0042-6989/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved. PII: S0042-6989(98)00223-5
1578
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
Magnocellular and parvocellular pathways show considerable functional overlap, however (Logothetis et al., 1990; Schiller, 1991) and complete isolation is never assured, but may be maximized by using carefully chosen spatial and temporal stimulus parameters (Murray & Kulikowski, 1983; Kulikowski et al., 1989; Kulikowski et al., 1991; Kulikowski, Robson & McKeefry, 1996). Low to medium spatial frequencies allow both systems to respond to chromatic and achromatic stimuli, while higher spatial frequencies bias responses to achromatic stimuli towards the parvocellular channel, and very low spatial frequencies bias the response towards the magnocellular channel (Kulikowski et al. 1989). Moderate to high spatial frequency stimuli with a large number of spatial cycles (more than eight to ten for red-green stimuli) may introduce chromatic aberration, with the result that only one of the two spatial sinusoids is brought into sharp focus (Flitcroft, 1989; Kulikowski, 1991). In addition, large stimulus fields may introduce variations in isoluminance with eccentricity (Zrenner, 1983). When low contrast stimuli are used, however, responses to large or small fields show similar characteristics (Kulikowski et al., 1991) suggesting that contrast control is more important than field size for the selective stimulation of magnocellular and parvocellular channels of the visual system. It has been noted that chromatic cells respond poorly to chromatic stimulus offset (Jankov, 1978; Zrenner, 1983). The pattern-offset response to chromatic stimuli appears to be present when moderate contrast (0.44) is used, and diminished under lower (0.22) contrast conditions (Kulikowski et al., 1991) (see their Fig. 2). For achromatic stimuli, the response to pattern-offset diminishes with increasing spatial frequency, suggesting that the offset response is weakened when elicited from
P-cells, responsive to high spatial frequency patterned stimuli (Murray & Kulikowski, 1983). In the present study, transient visual evoked responses to high contrast luminance-modulated and chromatic stimuli of low spatial and temporal frequency were recorded from four human adults, all aged under 35 years: three colour-normal females and one protanomalous male. The principal aim of the study was to observe characteristics of the evoked response to large field, high contrast chromatic stimulation as luminance differences are introduced (high contrast and large field stimuli were used with the intention of comparison with responses to similar stimuli recorded in infants, not reported here) (Suttle, Anderson & Harding, 1997). Our results demonstrate a lack of response to pattern-offset of isoluminant chromatic stimuli, a strong response to pattern-offset of luminance-modulated stimuli, and distinct morphology of the response to onset of luminance and chromatic patterns. These findings, with non-optimal stimuli, appear to reflect the characteristics of underlying physiological mechanisms, with luminance-responsive cells detecting contrast change, and chromatic-sensitive cells responding to the absolute level of contrast. 2. Methods
2.1. Subjects Three female (AE, FF and VT) and one male (IP) adult were included in the study, age range 26–34 to avoid age-related changes in ocular media density (Boettner & Wolter, 1962). For each subject, colour vision was tested using the Ishihara pseudoisochromatic plates and the Farnsworth–Munsell 100-Hue test under a standard illuminant C. The 100-Hue and Ishihara test indicated protanomaly for this subject (Farnsworth, 1943; Kinnear, 1970). The three female subjects all demonstrated normal colour vision on both tests.
2.2. Stimuli
Fig. 1. Colour luminance ratios representing individual isoluminant points for three colour-normal observers (FF, AE and VT) and one protanomalous observer (IP; red-green match only carried out in this case), determined by heterochromatic flicker photometry at a temporal frequency of 20 Hz. Photometric isoluminance is at a ratio of 0.5.
, tritan matches; , red-green matches.
Horizontal, luminance-modulated and chromatic (green–purple or red–green) sinusoidal gratings of spatial frequency 1.0 cpd were generated using a VSG 2/2 graphics board (Cambridge Research Systems) with 14-bit DACs. Low spatial frequency was chosen to avoid chromatic aberrations. Gratings were displayed on a 17-in. diameter CRT colour monitor (Eizo flexscan 562-T) at a non-interlaced frame rate of 150 Hz and line scan rate of 66 kHz. The resolution of the screen was 380 lines× 800 pixels. The monitor was viewed binocularly at a distance of 1 m, which allowed 14 cycles of sinusoid to be displayed. The angular subtense of the display was 18.5° horizontally×14° vertically.
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
1579
Fig. 2. Visual evoked responses to red-green (A) and tritan (B) stimuli at the full range of colour luminance ratios recorded from colour-normal adult VT. The points of pattern onset and offset are indicated by arrows on the abscissa.
The red, green and blue colour guns of the monitor were gamma-corrected, to ensure that luminance output was proportional to voltage input, using Cambridge Research Systems’ VSG software. Calibration studies showed that the phosphor chromaticity co-ordinates for each gun remain stable for at least 2 h, beginning 10 min after switching the monitor on. Recordings were always completed well within this time window. Chromatic stimuli were presented along two axes in CIE colour space: a tritanopic confusion axis intended to stimulate selectively the S-cone pathway, and an axis orthogonal to this to stimulate the L – M pathway. Both axes passed through a white point at x = y =0.31. Chromaticity co-ordinates were x= 0.36, y = 0.45 (greenish–yellow) and x =0.27, y= 0.22 (purple) for the tritan stimulus, and x =0.38, y =0.275 (red) and
x= 0.23, y= 0.35 (green) for the red-green stimulus. Chromaticity co-ordinates were measured using the Bentham PMC 3B spectroradiometer (Bentham Instruments, UK). Isoluminance of each colour scheme was determined in each of the three colour-normal adults, using heterochromatic flicker photometry. Red–green or green– purple stimuli were phase-reversed at a temporal frequency of 20 Hz. Isoluminance was similar for each colour-normal adult, and closely approximated photometric isoluminance (see Fig. 1). For this reason, stimuli were presented to each adult at a range of colour luminance ratios centered about photometric isoluminance (ratio 0.5). Colour luminance ratio values used here indicate the ratio of green luminance to total (green+ purple) luminance in the tritan stimulus, and
1580
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
Fig. 2. (Continued)
red luminance to total (red +green) luminance in the red – green stimulus. The ratios used were 0.0, 0.2, 0.4, 0.45, 0.5, 0.55, 0.6, 0.8 and 1.0. Thus, over the full range of colour luminance ratios, the red-green stimulus, e.g. would change from green – black through red– green to red–black. The mean luminance of the display was maintained at 20 cpd/m2, for all stimuli, and was measured using a Minolta LS 110 photometer. For the tritan stimulus at photometric isoluminance, cone contrast was 0.0, 0.0 and 0.82 for L- M- and S-cones, respectively. For the isoluminant red-green stimulus, these values were 0.15, 0.28 and 0.04, respectively (Cole & Hine, 1992). These stimuli were at the maximum contrast level available using our monitor; VEP threshold is at approximately 6 and 9% of these levels (Rabin et al., 1994). Michelson contrast of the luminance-modulated stimuli (colour ratio 0.0 or 1.0) was 0.9.
Abrupt pattern-onset presentation was used at a temporal frequency of 2 Hz, with a stimulus duration of 100 ms and inter stimulus interval of 400 ms, during which the stimulus was replaced by a blank screen of the same mean chromaticity and luminance. This temporal duty cycle was chosen to maximize the response amplitude to onset-offset stimuli (Rabin et al., 1994). Onset-offset presentation was chosen in preference to pattern-reversal to allow a clear morphological distinction between responses to chromatic and luminance stimuli (Murray et al., 1987; Berninger et al., 1989; Kulikowski et al., 1989; Rabin et al., 1994). Stimuli were presented at the full range of colour ratios in pseudo-random order. Responses were recorded to both types of chromatic stimuli from the three colournormal subjects, and to only the red-green stimuli from the colour-defective subject (IP).
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
2.3. VEP recording Transient visual evoked responses were recorded from Oz of the International 10 – 20 system, referred to Cz and Fz was used as an earth. Signals were bandpass filtered (1–50 Hz) and 50 sweeps were averaged using the Medelec Sapphire 4E. Silver-silver chloride 1 cm diameter electrodes were used. Omniprep skin preparation was used on each electrode position before fixing with Blenderm tape. Resistances were always below 8 kV.
3. Results Fig. 2 shows visual evoked responses to tritan and red-green stimuli at the full range of colour luminance ratios for colour-normal subject VT. The responses shown here are typical of responses recorded from all three colour-normal subjects. At photometric isoluminance (colour ratio 0.5) the peak response component is a negativity occurring approximately 150 ms or 120 ms after stimulus onset for tritan and red-green, respectively, followed by a smaller positivity (around 100 ms later, at 220–240 ms after stimulus onset) in agreement with previous findings (Berninger et al., 1989; Rabin et al., 1994). As luminance differences are introduced to the chromatic stimulus, a second negative component occurs on the descending limb of the onset response, at a latency of approximately 170 ms for both types of chromatic stimuli. This second negativity is not apparent at isoluminance, and becomes better defined with luminance difference. Concurrent with this, the earlier negativity, which is the peak response component at isoluminance, becomes diminished as luminance differences are introduced to the stimulus. At colour ratios of 0.0 and 1.0 (luminance-modulated stimuli) this earlier negative component is only vaguely apparent, and is slightly later than the peak response to luminance stimulus onset, a small positivity at 80 – 100 ms latency. The clearest component in the luminance response (colour ratios 0.0 and 1.0) is the response to stimulus offset, a P-N-P component, the negativity of which occurs approximately 130 ms after offset. Fig. 3 shows the evoked response to the luminance-modulated stimulus (colour ratio 0.0) with offset having occurred at 100 ms, as for all other responses here, and in the lower trace with offset at 200 ms. The P-N-P component occurs correspondingly later, confirming that this is a response to stimulus offset. The response to stimulus offset also demonstrates changing morphology as luminance ratio of the chromatic stimulus is varied. At isoluminance, the response to red-green stimuli shows a very small P-N-P component, the negativity of which occurs approximately 230 ms after stimulus offset. As luminance differences are
1581
introduced, this component increases in amplitude and definition until at colour ratios of 0.0 and 1.0 the response to stimulus offset is the clearest component in the evoked response. The tritan response also shows this morphological change, with a response to offset becoming clearer at colour ratios away from isoluminance, although the trend is less clear than in the red-green responses. As demonstrated by Fig. 2, the evoked responses to stimulus onset and offset show opposing morphological changes towards isoluminance. The onset response becomes better defined, and the offset response diminishes, as isoluminance is approached. Fig. 4 illustrates this trend in terms of onset and offset response amplitudes, at the full range of colour luminance ratios, for all three colour-normal observers. Evoked responses recorded from protanomalous subject IP to red-green stimuli are shown in Fig. 5. At a colour luminance ratio of 0.6 the evoked response shows no clear components. This ratio is close to red–green isoluminance for this subject, as determined by heterochromatic flicker photometry (see Fig. 1). At all other ratios, a small positivity is seen at approximately 130 ms after stimulus onset. In addition, the P-N-P response to stimulus offset is apparent 120–130 ms after stimulus offset for responses to colour ratios other than 0.6.
4. Discussion When onset–offset presentation is used, peak responses to luminance and chromatic pattern-onset differ in polarity indicating that responses to luminance and chromatic pattern onset are dominated by, but not
Fig. 3. The visual evoked response to stimulus offset is demonstrated here by doubling the onset duration to 200 ms, so that stimulus offset occurs 100 ms later in the lower than in the upper trace (colour-normal adult FF). Onset and offset times are indicated on the abscissa.
1582
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
Fig. 4. Amplitude of the evoked potential to stimulus onset (A) and offset (B) is shown for all three colour-normal subjects, at the full range of colour luminance ratios. In (B), points at zero indicate that no component could be distinguished from the response. Peak to peak amplitudes were measured from the peak of the negative component to the following positivity of the onset response, and from positive peak to negativity of the offset response. The figure legend relates symbols to subjects, identified by their initials.
elicited exclusively from, distinct groups of cortical cells (see Section 1). In the present study, the peak response to chromatic onset stimulation was found to be negative-going, and the peak response to luminance onset was a positivity, in agreement with previous findings (Murray et al., 1987). While these responses are unlikely to be entirely selective, response morphology suggests that the responses to chromatic stimulation at
photometric isoluminance are dominated by the activity of colour-selective cortical cells, and that responses to luminance modulation are dominated by broad-band (luminance-responsive) cells. As luminance differences were introduced to the chromatic stimulus, a second negativity appeared in the peak response to chromatic pattern-onset, being absent or minimal at isoluminance and becoming increasingly
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
well-defined at colour ratios away from isoluminance (see Fig. 2). This morphological change suggests that the second component, elicited as luminance differences were introduced to the stimulus, is a response not to chromatic contrast but to luminance modulation. A similar change in morphology has been observed previously in evoked responses to chromatic stimuli (Kulikowski et al., 1989). In that study, the peak response to chromatic pattern onset was found to consist of two negative components, the earlier of which was found to be related to chromatic contrast, and the second to pattern elements of the stimulus, because this component increased in amplitude as the stimulus changed from one- to two-dimensions, and because a similar component was present in the achromatic response. The evoked response to stimulus offset is minimal at photometric isoluminance and becomes increasingly well-defined as luminance differences are introduced to the stimulus. Chromatic cells respond weakly to stimulus offset (Jankov, 1978; Zrenner, 1983), responding to absolute contrast rather than to contrast change. The contrast of the stimulus at offset is zero, and hence the response to this stimulus from chromatic cells is ex-
Fig. 5. Visual evoked responses to red-green stimuli at the full range of colour luminance ratios are shown, recorded from protanomalous adult IP. Pattern onset and offset times are indicated on the abscissa. The red-green isoluminant point for this observer determined by heterochromatic flicker photometry was 0.55.
1583
pected to be zero. Luminance-responsive cells respond to contrast change, and so these cells respond to both onset and offset phases of the stimulus. These response patterns are reflected in the evoked responses recorded in the present study, with the response to isoluminant chromatic stimulation showing minimal offset response, and the response to isochromatic luminance stimulation showing maximal offset response. At colour luminance ratios either side of photometric isoluminance, the offset response begins to appear, and becomes more well defined with increasing luminance difference. The small response to stimulus offset seen in the evoked response to isoluminant red–green stimulation (Fig. 2) may reflect the fact that the stimulus was at photometric isoluminance, a slight deviation from the individual’s isoluminant point (see Fig. 1). Visual evoked responses have been widely used to investigate colour vision deficiencies. In general, a stimulus along an appropriate colour confusion axis elicits diminished or absent evoked responses from a dichromatic or anomalous trichromatic observer (Kinney & McKay, 1974; Regan & Spekreijse, 1974; Crognale, Switkes, Rabin & Schneck, 1993). In the present study, the evoked response of a protanomalous observer to red–green stimulus pattern-onset is unclear at all colour luminance ratios, and the response to patternoffset is absent at a colour luminance ratio of 0.6 (see Fig. 5), close to this observer’s isoluminant point (Fig. 1). At ratios either side of this, the evoked response to stimulus offset appears and becomes more defined at ratios away from this point, presumably reflecting the increasing activity of luminance-responsive cells. At a R/R + G colour luminance ratio of 0.6, the red grating would appear brighter than the green to a normal trichromatic observer, but is close to perceived isoluminance for a protanomalous observer (Allen, Banks & Norcia, 1993). At this point the red and green components of such a stimulus, designed to modulate L- and M-cone types, provide equal stimulation of the remaining functional cone type being stimulated. Using the transient visual evoked response to low spatial and temporal frequency tritan and red-green chromatic stimuli at a range of colour luminance ratios spanning photometric isoluminance, we have demonstrated morphological changes to the onset and offset response components as luminance differences are introduced to the chromatic stimulus. Such changes to the onset response with colour luminance ratio are well known, and have been used to demonstrate the involvement of magnocellular and parvocellular pathways in the evoked response (Murray et al., 1987). It is also well understood that chromatic cells respond poorly to stimulus offset (Jankov, 1978; Zrenner, 1983). The results of the present study show that at photometric isoluminance the response to stimulus offset is minimal or absent, and that at colour luminance ratios removed
C.M. Suttle, G.F.A. Harding / Vision Research 39 (1999) 1577–1584
1584
from isoluminance this response component begins to appear, and becomes more apparent and well defined with increasing luminance difference. In one protanomalous observer, the response to chromatic contrast was found to be minimal, and the response to pattern-offset showed increasing definition with luminance difference, as also seen in the responses from colour-normal observers. These results demonstrate that morphological changes to the response component to pattern-offset may be used in addition to pattern-onset response characteristics, in colour-normal and colour-defective observers, as part of the signature of the transient evoked response to chromatic stimuli at isoluminance.
Acknowledgements CMS thanks the ISCEV Board of Directors for a grant to travel to the 1996 annual symposium, where this work was originally presented.
References Allen, D., Banks, M. S., & Norcia, A. M. (1993). Does chromatic sensitivity develop more slowly than luminance sensitivity? Vision Research, 33, 2553 – 2562. Anderson, S. J., Holliday, I. E., Singh, K. D., & Harding, G. F. (1996). Localization and functional analysis of human cortical area V5 using magneto-encephalography. Proceedings of the Royal Society of London. Series B: Biological Sciences, 263, 423 – 431. Baseler, H. A., & Sutter, E. E. (1997). M and P components of the VEP and their visual field distribution. Vision Research, 37, 675 – 690. Berninger, T. A., Arden, G. B., Hogg, C. R., & Frumkes, T. (1989). Separable evoked retinal and cortical potentials from each major visual pathway: preliminary results. British Journal of Ophthalmology, 73, 502 – 511. Boettner, E. A., & Wolter, J. R. (1962). Transmission of the ocular media. In6estigati6e Ophthalmology, 1, 776–783. Cole, G. R., & Hine, T. (1992). Computation of cone contrasts for color vision research, behavior research methods. Instruments and Computers, 24, 22 – 27. Crognale, M., Switkes, E., Rabin, J., & Schneck, M. E. (1993). Application of the spatio-chromatic visual evoked potential to detection of congenital and acquired colour deficiencies. Journal of the Optical Society of America A, 10, 1818–1825. Farnsworth, D. (1943). The Farnsworth-Munsell 100-Hue and dichotomous tests for color vision. Journal of the Optical Society of America, 33, 568 – 578.
.
Flitcroft, D. I. (1989). The interactions between chromatic aberration, defocus and stimulus chromaticity: implications for visual physiology and colorimetry. Vision Research, 29, 349 – 360. Jankov, E. (1978). Spectral sensitivity of off-response in human VECP during selective chromatic adaptation (author’s translation). Albrecht 6on Graefes Archi6 fur Klinische und Experimentelle Ophthalmologie, 206, 121 – 133. Kinnear, P. R. (1970). Proposals for scoring and assessing the 100Hue test. Vision Research, 10, 423 – 433. Kinney, J. A., & McKay, C. L. (1974). Test of color-defective vision using the visual evoked response. Journal of the Optical Society of America A, 64, 1244 – 1250. Kulikowski, J. J., Murray, I. J., & Parry, N. R. A. (1989). Electrophysiological correlates of chromatic-opponent and achromatic stimulation in man. In B. Drum, & E. Verriest, Colour 6ision deficiencies IX. Dordrecht: Academic, 145 – 153. Kulikowski, J. J., Murray, I. J., & Russell, M. H. A. (1991). Effect of stimulus size on chromatic and achromatic VEPs. In B. Drum, J. D. Moreland, & A. Serra, Colour 6ision deficiencies X. Dordrecht: Academic, 51 – 56. Kulikowski, J. J. (1991). On the nature of visual evoked potentials, unit responses and psychophysics. In A. Valberg, & B. B. Lee, From pigments to perception. New York: Plenum, 197 – 208. Kulikowski, J. J., Robson, A. G., & McKeefry, D. J. (1996). Specificity and selectivity of chromatic visual evoked potentials. Vision Research, 36, 3397 – 3401. Logothetis, N. K., Schiller, P. H., Charles, E. R., & Hurlbert, A. C. (1990). Perceptual deficits and the activity of the color-opponent and broad-band pathways at isoluminance. Science, 247, 214– 217. McKeefry, D. J., Russell, M. H., Murray, I. J., & Kulikowski, J. J. (1996). Amplitude and phase variations of harmonic components in human achromatic and chromatic visual evoked potentials. Visual Neuroscience, 13 (4), 639 – 653. Murray, I. J., & Kulikowski, J. J. (1983). VEPs and contrast. Vision Research, 23, 1741 – 1743. Murray, I. J., Parry, N. R. A., Carden, D., & Kulikowski, J. J. (1987). Human visual evoked potentials to chromatic and achromatic gratings. Clinical Vision Sciences, 1, 231 – 244. Rabin, J., Switkes, E., Crognale, M., Schneck, M. E., & Adams, A. J. (1994). Visual evoked potentials in three-dimensional color space: correlates of spatio-chromatic processing. Vision Research, 34, 2657 – 2671. Regan, D., & Spekreijse, H. (1974). Evoked potential indications of colour blindness. Vision Research, 14, 89 – 96. Schiller, P. H., Logothetis, N. K., & Charles, E. R. (1990). Functions of the color-opponent and broad-band channels of the visual system. Nature, 343, 68 – 70. Schiller, P. H. (1991). The color-opponent and broad-band channels of the primate visual system. In A. Valberg, & B. B. Lee, From pigments to perception. New York: Plenum Press. Suttle, C. M., Anderson, S. J., & Harding, G. F. A. (1997). A longitudinal study of visual evoked responses to tritan stimuli in human infants. Optometry and Vision Science, 74 (9), 717–725. Zrenner, E. (1983). Neurophysiological aspects of color 6ision in primates. Berlin: Springer-Verlag Chapter 6.