Behavioral and physiological study of color processing in human brain. 

Hitoshi Sasaki, Sumie Matsuura1, Akiko Morimoto and Yutaka Fukuda

Department of Physiology and Biosignaling, Osaka University Graduate School of Medicine, 2-2 Yamadaoka,Ssuita Japan and 1San Cuore, 4-22-2 Nakayamate St., Chuo-ku, Kobe, Japan.
Key words; simple reaction time, discrimination time, alpha-blocking, color detection, cerebral hemispheric dominance
running title; color processing in human brain 
address to which proofs should be sent; 
  Department of Physiology and Biosignaling, (A5)
  Osaka University Graduate School of Medicine,
  2-2 Yamadaoka, Suita 565-0871 Japan 
  fax number; +81-6-6879-3619

1410 Words

    Although color processing is one of  the important features in visual information processing in humans, little is known about the color processing in the human brain. In the present study we asked two questions using human subjects. First, how color stimulation of different hue (wavelength), is processed within the brain. Secondly, we asked whether or not hemispheric difference exists between right and left hemispheres in processing of color stimulation.  To answer these questions we presented chromatic stimuli,  matched photometrically in luminance,  on achromatic background or achromatic circle of same size with the same luminance�@ and measured simple and discriminative reaction times (RTs) as well as latencies of alpha-blockings in EEG induced by these color stimuli. 

1) Effects of hue on simple reaction time
RT is a minimal time required for detection of sensory input and the time necessary for the motor output. It has been widely used to evaluate brain functions including visual information processing [1]. Although many studies have been reported as to the effect of hue on simple RT, the consistent result has not been reported for the difference among different hue stimuli  [2-4]. In the present experiments, subjects were seven male undergraduate students (aged 20 - 23) with normal or corrected normal vision. They  were instructed to press a button as fast as possible after they perceive the visual target. All visual stimuli were presented on a CRT display with an uniform achromatic background of 24 cd/m2  (10��x10��). As a fixation point, an achromatic 2 deg-circle of 10 cd/m2 was always presented at the center of the display. Target stimuli were 2 deg-circles with four different main wavelength (R: 635 nm, Y: 580 nm, G: 548 nm, B: 463 nm). Luminance of these stimuli was matched to the achromatic circle of 10 cd/m2 by flicker photometry. Saturation was adjusted to blue stimulus of 60 %. With a random delay of 3-5 s, following a ready signal (tone, 55 dB, with a background level of 42 dB, SPL), one of the target stimulus was presented on the fixation point with an inter-trial interval (ITI) of 15 s (10-20s). As shown in Fig. 1a, simple RT to  blue was significantly longer than that to red.  Similar results were obtained in another experiment using different main wavelength (R: 600 nm, Y: 577 nm, G: 537 nm, B: 465 nm, Fig. 1b). Here again, the RT to blue was significantly longer than those to red, yellow and green. 

2) Effects of hue on EEG alpha-blocking
While alpha waves in EEG appear dominantly in resting state, its amplitude  showed marked reduction in response to visual stimulation (Fig 2a). This phenomenon is called as alpha-blocking, and can be used as a sign of changes in arousal level of the brain. Ten undergraduate students (aged 21-23) with normal or corrected normal vision faced to the CRT with a background of  10cd/m2 . One of the three chromatic stimuli �iR: 570 nm, G: 535 nm, B: 445 nm, �T��x�T���jwas presented for 2 sec with an ITI of at least 10 s. Bipolar EEG was recorded from the occipital cortex (TC 0.3 s, Hi-cut at 60 Hz) together with the EOG (TC 5 s, Hi-cut at 60 Hz). Five out of 10 subjects showed continuous alpha waves in EEG recordings even under eye-opened state (Fig.2a). These subjects showed a distinct alpha-blocking to chromatic stimuli without any changes in luminance and the latency was measured for each hue stimulation.  As shown in Fig.2b, mean latency of alpha-blocking was longer to blue as compared to those to red and green. This result is consistent with the above result that the RT was longer to blue stimulation. These results are consistent with a recent finding in monkeys that in the brain blue-on information is processed via a koniocellular pathway  which has slower conduction velocities than those devoted for the processing of red and green [5]. 

3) Effects of hue on discriminative reaction time
 Discriminative RT was measured in the discrimination of one color stimulus from the others with a go-no go task, so that it includes the time required for discrimination of hue in addition to its detection time which is mainly analyzed in the case of simple RT.  Six undergraduate subjects were asked to press a button only to a target, which was previously instructed, out of three chromatic stimuli (R: 635 nm, G: 548 nm, B: 463 nm). Discriminative RT was longer than simple RT in all of the chromatic stimuli. Discriminative RT was significantly longer to blue (Fig. 3a). Moreover, difference of time calculated by subtracting simple RT from the discriminative RT in each hue, was also longer to blue (R: 65.0�}15.3,  G: 86.1�}9.6,  B: 93.3�}14.6). A nearly significant and a strong tendency of difference was observed between the times in blue and those in red (t(5)=1.787, p=.134). Theses results suggest that not only detection but also discrimination process for blue takes longer.

4) Right hemispheric dominance in color detection
It is generally known that right hemisphere is dominant in visuospatial information processing [6]. In the present study we examined whether or not hemispheric dominance exists in color detection. Human optic nerves originating from the nasal retina project to the contralateral visual cortex, while the nerves from the temporal retina project to the ipsilateral visual cortex. Pyramidal tracts from the motor cortex control  contralatelal hand movements, thus right motor cortex innervate left hand and vice versa. Based upon these double crossed projections, we evaluated the time required for detection of color in each hemisphere by comparing RTs to visual stimuli presented either right or left visual filed and responded by the ipsilateral hand [7]. Ten undergraduate students with normal vision (ages ranged 21-23, both gender, right-handed) were asked to attend  in two blocks of experiments. The target , either chromatic (R 635 nm, G 535 nm, B 445 nm) or achromatic (12-20 cd/m2), of  2 deg-circle was presented on a background of 10 cd/m2. EOG was recorded to monitor the eye fixation. In the first experiment, the target was presented at the center of visual field and there was no significant difference between RTs by right and left hand. In the second one, the target appeared at 4 deg horizontally from the fixation point in either right or left visual field. As shown in Fig. 3b, RT by the left hand to the chromatic target presented in the left visual field was significantly shorter than those by the right hand to the target in the right visual field. On the other hand, there was no significant difference between the RTs to achromatic targets. These data show that the right hemisphere is dominant in detection of color stimulus in human subjects.

Conclusion
 Behavioral study using RT showed that blue information is uniquely processed in the brain.  Simple RT was longer to blue as compared to red, yellow or green. Results of EEG response were consistent with this finding. 
Longer time for processing of blue information seems, at least�@partly, to be ascribed to longer  processing time in the retina and  subcortical pathway. 
Results of discriminative RT show that discrimination time for blue is also longer. As the discrimination process belongs to one of function of the color center, these findings suggest that processing of blue in the cortical visual center also takes a longer time. A comparison of reaction time between  right and left hemispheres  shows that the right hemisphere is dominant in detection of color.
 
References
[1] Donders  FC (1969)  On the speed of mental processes.  Acta Psychol. (Amst)., 30:412-431.
[2] Schwartz SH  (1995) Dependence of visual latency on wavelength: predictions of a neural counting model. J. Opt. Soc. Am. A., 12:2089-2093.  
[3] Finn  JP, Lit A. (1971) Effect of photometrically matched wavelength on simple reaction time at scotopic and photopic levels of illumination.  Proceed. Annu Conv. Amer. Psychol. Assoc., 6 (Pt.1): 5-6. 
[4] Ueno T, Pokorny J, Smith VC (1985) Reaction times to chromatic stimuli. Vision Res., 25:1623-1627. 
[5] Martin PR, White  AJ, Goodchild  AK et al. (1997) Evidence that blue-on cells are part of the third geniculocortical pathway in primates. Eur. J. Neurosci., 7:1536-1541.
[6] Nebes RD (1973) Perception of spatial relationshis by the right and left hemispheres in commissurotomized man. Neuropsychologia, 11:285-289. 
[7] Berlucchi G, Heron W, Hyman R et al. (1971) Simple reaction times of ipsilateral and contralateral hand to lateralized visual stimuli. Brain, 94:419-430.

 Figure Legends
Fig. 1a,b. Effects of hue on simple reaction time.  a  Simple RT to chromatic stimuli (R: 635 nm, Y: 580 nm, G: 548 nm, B: 463 nm) in 7 subjects. There was a significant difference between RTs to red and blue  (t(6) = 2.885, p<.05).  Mean with SD.  b  Simple RT to chromatic stimuli with different main wavelength (R: 600 nm, Y: 577 nm, G: 537 nm, B: 465 nm) in 10 different subjects. RT to blue was significantly longer than any other hue (*: p<.05, ** : p<.01). Mean with SD.  
Fig. 2a,b. EEG alpha-blocking by chromatic stimuli.  a  In a half of subjects (5/10) , continuous alpha waves were observed even under eye-opened state and distinct alpha-blocking was elicited by chromatic stimulus without any change in luminance.  b Mean latency of alpha-blocking by blue was significantly longer than that by red�it(37)=2.274, p<.05�j. Mean with SE.
Fig. 3a. Discriminative RT for chromatic stimuli was significantly longer to blue than to red (t(5)=3.086, p<.05). Mean with SE.  b  RT by left hand to target in left visual field (LL) was significantly shorter than RT by right hand to target in right visual field (RR) only when the target was chromatic (t(9) = 3.171, p<.05) . Mean with SE.