Chapter 39 Neurophysiologic tools to explore visual cognition

Advances in Clinical Neurophysiology (Supplements to Clinical Neurophysiology Vo!. 54) Editors: R.C. Reisin, M.R. Nuwer, M. Hallett, C. Medina (02002 Elsevier Science B.V. An rights reserved.

261

Chapter 39

Neurophysiologic tools to explore visual cognition Shozo Tobimatsu Department of Clinical Neurophysiology, Neurological Institute, Faculty ofMedicine, Graduate School of Medical Sciences, Kyushu University, Fukuoka 812-8582 (Japan)

Introduction The visual system analyzes the spatial, temporal and chromatic aspects of objects via the multiple, parallel channels . In primates, visual information is processed via two parallel visual pathways; the parvocellular (P) and magnocellular (M) pathways (Livingstone and Hubel 1988; Zeki et al. 199 I; Celesia and DeMarco 1994). The P-pathway is thought to be responsible for detecting form and color because of its high spatial resolution, color sensitivity, low contrast sensitivity, and slow temporal resolution. The use of onset-offset mode rather than contrast reversal stimulus is more appropriate for the activation of the P-system (Murray et al. 1987). The M-system is considered to be responsible for detecting dynamic form and motion because of its fast temporal resolution, high contrast sensitivity, color insensitivity, and low spatial resolution. Several lines of evidence suggest that the two systems exist in humans (Livingstone and Hubel 1988; Zeki et al. 1991; Zeki 1993). We have developed the visual stimuli designed to

* Correspondence to: Dr. S. Tobimatsu, Department of Clinical Neurophysiology, Neurological Institute, Faculty of Medicine, Graduate School of Medical Sciences, Kyushu University, Fukuoka 812-8582, Japan. Fax: +81-92-642-5545. E-mail: [email protected]

preferentially stimulate each pathway and carried out electrophysiological studies in humans (Tobimatsu et al. 1995, 1999; Tobimatsu and Kato 1998; Arakawa et al. 1999). We report here recent our findings on visual evoked potentials (VEPs), visual evoked magnetic fields (VEFs) and visual event related potentials (ERPs) with regard to visual cognition.

Neural generators of pattern reversal visual evoked potentials The pattern reversal YEP is a pertinent tool for assessing the visual function, however, its neural generators have remained unknown. We had an opportunity to record optic tract responses in Parkinsonian patients during stereotactic posteroventral pallidotomy (Tobimatsu et aI. 1997). Pallidal VEPs showed an initial positive deflection (P50) followed by a negativity (N80; Fig. IA). P50 and N80 were near-field potentials because they were limited to the vicinity of the optic tract (Tobimatsu et al. 1997). In addition, we estimated the location of equivalent current dipole (ECD) of pattern reversal VEPs in healthy subjects by using magnetoencephalography (MEG; Shigeto et al. 1998). VEFs showed three peaks ofN75m, PI00m and Nl45m that corresponded in time to N75, PIOO and Nl45

262

A

C

Pallidal and Scalp VEPs

Chromatic pattern

Nao

GPi

\~ pso

Scalp

•

YEP

0/~~'

VEF

"'T

N145

N~

~~,.J; \

N120

.-.......1\

.

14

11V

P100

2 11V 1

N120m

100 IT

L-

0

100

B

200

0

300ms

D

Pattern Reversal N75

YEP

100

200ms

Apparent Motion

N145

~'-'/~-V""'--_/--

YEP

V-V---. P120

P100 N7Sm

VEF

~ P100m

2 11V 1

VEF

2 11V P120m

100 IT

1

o

Pattern reversal • Apparentmotion

100 IT

L-

0

100

200

300ms

0

100

200ms

Fig. I. (A) Pallidal and scalp VEPs to pattern reversal stimulation. Pallidal VEPs showed P50 with an onset latency about 30 ms and N80; both were limited to the ventral part of the globus pallidus internus (GPi). Scalp VEPs consisted ofN75, PIOO and N145. (B) Simultaneous recording ofVEPs and VEFs to pattern reversal stimulation. The responses of37 channels were superimposed in VEFs. The latencies of each component were delayed about 20 ms due to low luminance of the visual stimuli in a magnetically shielded room in (B-O). (C) Simultaneous recording of VEPs and VEFs to the onset of red-green isoluminant chromatic pattern. The source ofNl20 was estimated to be in the primary visual cortex. (0) Simultaneous recording ofVEPs and VEFs to apparent motion display. The source ofP 120 was estimated in the primary visual cortex, which was close to the source ofPIOO. (From Tobimatsu et al. 2000, with pennission.)

in VEPs (Fig. IB). The ECDs of N75m, PlOOm and N 145m were located in the primary visual cortex (V 1) contralateral to the stimulated visual field. The direction ofthe current flow ofECDs ofN75m and N145m was from the medial to the lateral aspect of the head, whereas that for PI00m was directed mesially when viewed in a coronal section. It is assumed that P50 and N80 of intracranial pattern reversal VEPs reflect the compound action potentials of the optic tract (Tobimatsu et al. 1997). Our results, however, indicate that N75 may represent an initial response of V I and that they are

not generated in the optic tract. N75 and PI 00 may be generated from different neuronal populations in VI. Our study further suggests that N145 may also be generated in VI.

Visual evoked potentials and magnetic fields to color and motion stimulation Although there are overlapping portions of visible spatial and temporal frequencies between the P- and M-systems, there is a general agreement

263

that M outputs dominate at low contrasts and higher temporal frequencies while P outputs are predominant at higher spatial frequencies and chromatic contrasts (Murray et al. 1987). We hypothesize that the use of isoluminant red-green gratings at a low temporal frequency may preferentially stimulate the Pvsystem while the M-system is preferentially activated by motion stimulation (Tobimatsu et al. 1995). The visual stimuli were generated on a color CRT.The red and green phosphors were modulated sinusoidally in anti-phase to produce soluminant chromatic(red-green) sinusoidalgratings (Tobimatsu et al. 1995). High contrast (90%) black-and-white sinusoidal gratings were also used. The mean luminance ofboth the chromatic and achromatic gratings was 16 cd/m". Patterns appeared for 200 ms and were replaced by a yellowish white or grayish white background for 800 ms. Thus, the temporal frequency was 1 Hz. Nine gratings, spatial frequencies ranging from 0.5-8.0 c/deg were tested. An apparent motion (AM) display (Tobimatsu et al. 1995) was generated on a color CRT: two squares at opposite comers of a hypothetical square were presented together for a certain duration. These squares were then extinguished and squares at the remaining two comers presented for the same duration. With this procedure repeated in a continuous cycle without any intervening blank fields, either vertical or horizontal motion can be perceived. The distance between the center of each square and a fixation point was 2° of arc and the squares themselves subtended 30 and 60 min of arc. The speed ofalternation was 500 ms. VEPs were recorded from a mid-occipital electrode (2.5 em above the inion). A total of 100 responses were averaged by a minicomputer with a bandpass 0.5-120 Hz. The appearance of an isoluminant chromatic pattern produced a major negative peak at around 120 ms (NI20) (Fig. lC). Both the amplitude and the latency ofN120 highly depended on the spatial frequency. The N 120 amplitude was largest for 2.0 c/deg and was smaller at both smaller and larger gratings (Tobimatsu et al. 1995). VEPs to achromatic gratings also showed a negative peak at around 95 ms (N95). In contrast with the responses to chromatic gratings, the Nl amplitude was maximal at either 5.3 c/deg or 8.0 c/deg. The N120 was absent

or delayed in patients with acquired color deficits (Tobimatsu and Kato 1998), indicating that the N 120 was related to color sense. When we compared achromatic (high contrast yellow-black checks; 81%) and chromatic (low contrast red-green checks; 12%) VEFs (Tobimatus et al. 1999), two sources were located in V 1 and very close each other (Fig. IC). However, the latency of achromatic VEFs was about 20 ms shorter than that of chromatic VEFs. The magnitude of magnetic fields of N120 were greater than that of achromatic stimulation. In addition, low-contrast yellow-black checks (21%) dramatically reduced the VEF response. These findings indicate that the N120 reflects color information rather than contrast information. Therefore, VEPs to chromatic stimulation allow us to evaluate the function of the P-system in VI. VEPs to AM showed a major positive peak at around 120 ms (PI20; Fig. ID). AM display has been considered to preferentially activate the Msystem (Livingstone and Hube11988) and the subjects in our study perceived motion perception easily. The above observations indicate that characteristic potentials may distinguish between these two parallel visual systems in humans. VEFs were recorded to estimate the dipole sources of motion VEPs by using a 37-channel MEG (Tobimatsu et al. 1999, 2000). We compared pattern reversal VEFs with motion VEFs to AM. The sources of the PI 00 and P 120 were located in V 1 but the dipole ofP 120 was more lateral to that ofP 100 (Fig. 2B). Pattern reversal stimulation often produces a strong motion sensation. Spekreijse et al. (1985) proposed that pattern reversal VEPs consisted of a motion component but not contrast onset and offset responses. It is likely that the P120 as well as the PI 00 are related to motion sensation. Therefore, VEPs to AM display allow us to evaluate the function of the M-system in VI.

Event related potentials to color and motion discrimination We recorded ERPs to study the functions of the two streams of visual information mainly after VI (Arakawa et al. 1999). In a parvocellular task, 128

264

Parvocellular task 01

Freq.

Magnocellular task 01

N1~0(p)

N16,0(m)

Norm~

Norm~

SCD~

* SCD~

PD~

PD~

pz

Ram

pz ~4?0(m)

___

Norm~

Norm

SCD~

SCD~.

PD~ -200

a

200

400

600

800

Time (msec)

PD

~

~

**

----J'

~

+1

10 V IJ.

*It

1000

-200

a

200

400

600

800

1000

Time (msec)

Fig. 2. Grand averages of ERPs in each group. (Left panel) Response waveforms in the parvocelJular task. N160(p) at 01 and P400(p) at Pz are indicated by dotted lines. (Right panel) Responses in the magnocellular task. N 160(m) at 0 I and P400(m) at Pz are indicated by dotted line. In the SeD group, P400(p), N 160(m) and P400(m) were prolonged. In the PD group, P400(m) was prolonged witn normal N 160(m). Significant delayed peaks were indicated by asterisks. (From Arakawa et al. 1999, with perrnission.)

color dots with 5 mm of diameter were presented in a random spatial pattern on a uniform blue background. Dots appeared for 500 ms and disappeared for 2000 ms. Red dots were used as frequent stimuli (non-target) and rare stimuli (target) were green. Luminance of random dots and the background were equal. In a magnocellular task, a 3-D structured black virtual cylinder revolved clockwise at a fixed angular velocity of 60 deg/s whose surface was covered with 128 white random dots with 5 mm diameter. Each dot moved for 500 ms, and then was stationary for 2000 ms. The unstructured stimulation was generated by maintaining the velocity distribution while destroying the spatial relationship of the random dots each other. The 3-D structured stimuli were used as frequent stimuli, and the unstructured stimuli were used as rare stimuli. ERPs were recorded from 12 electrodes placed over the scalp with a bandpass between 0.05 and 500 Hz. In each session, 150 responses were stored and averaged offline.

In the P paradigm, frequent stimuli evoked a negative potential at around 160 ms (N 160(p)) which was maximal at the occipital region (Fig. 2). Grand averages ofERPs to rare stimuli evoked a large positive-going potential at around 400 ms (P400(p)) that was maximal at Pz (Fig. 2). In the M paradigm, all but one subject could immediately imagine that there were random dots on the surface of the rotating cylinder. Responses to frequent stimuli showed a large negative potential at around 160 ms (NI60(m)), which was observed exclusively at the temporal and occipital regions (Fig. 2). The rare stimuli evoked a large positive deflection at around 400 ms (P400(m)) that was maximal at Pz (Fig. 2). There were no significant differences in amplitude and latency between P400(p) and P400(m). However, a deuteranope lacked the P400(p) while P400(m) was normally evoked at 400 ms. This finding suggests that the two paradigms selectively activated each pathway, and that the P400(p) and P400(m) may reflect integrating processes of the visual information, particularly the neural proc-

265 esses of perception, cognition and differentiation of the visual stimuli used in the paradigms. It appears that this methodology has some advantage compared with the conventional YEP to investigate the visual processing streams after Vi. We have shown that the P- and M-pathways are affected in patients with spinocerebellar degeneration while the M-pathway is selectively impaired in Parkinson's disease (Fig. 2) (Arakawa et al. 1999). We have developed a technique of multimodality VEPs (Tobimatsu and Kato 1998) in which we recorded VEPs to 30' checkerboard patterns, R-G and B-W sinusoidal gratings and AM, and steadystate VEPs to B-W gratings in 15 normal controls and 14 patients to test the hypothesis that these potentials may derive from the functional subdivisions of the visual pathways. VEPs to 30' checks were abnormal in 10 eyes (7 patients); however, considering all five modalities, abnormal responses were seen in 20 eyes (12 patients). Abnormality rates were not equal among the visual stimuli, which thus suggested possible dysfunction of individual subdivisions in the visual pathways. We consider that use of multimodality VEPs may increase both understanding of the pathophysiology of the visual pathways and diagnostic yield. Furthermore, ERPs to color and motion discrimination may provide: the additional information mainly after Vi.

Conclusions Our study suggests that the functional roles of the P- and M-systems can be evaluated electrophysiologically in humans by using appropriate visual stimuli. Therefore, a combined use ofVEPs, VEFs and ERPs is useful for exploring the visual cognition.

Acknowledgments We thank Drs. K. Arakawa, K. Ishido, M. Kato, H. Shigeto and F. Shima for their valuable help. This study was supported in part by 'Grant-in-aid for General Scientific Research, from the Ministry of Education, Science and Culture, Japan.

References Arakawa, K.. Tobimatsu, S., Kato, M. and Kira, J. Parvocelluar and magnocellular visual processing in spinocerebellar degeneration and Parkinson's disease: an event-related potential study. Clin. Neurophysiol., 1999, 110: 1048-1057. Celesia, G.G and DeMarco, P. Anatomy and physiology of the visual system. 1. Clin. Neurophysiol., 1994, II: 482-492. Livingstone, M. and Hubel, D. Segregation of form color, movement and depth: anatomy, physiology, and perception. Science, 1988,240: 740-749. Murray, I.J., Parry, N.R.A., Carden, D. and Kulikowski, U. Human visual evoked potentials to chromatic and achromatic gratings. Clin. Vision Sci., 1987, 1: 231-244. Shigeto, H., Tobimatsu, S., Yamamoto, T., Kobayashi, T. and Kato, M. Visual evoked cortical magnetic responses to checkerboard pattern reversal stimulation: a study on the neural generators of N75, P100 and N145. 1. Neurol. Sci., 1998, 156: 186-194. Spekreijse, H., Maier, DJ. and Regan, D. Flicker and movement constituents of the pattern reversal response. Vision Res., 1985, 25: 1297-1304. Tobimatsu, S. and Kato, M. Multimodality visual evoked potentials in evaluating visual dysfunction in optic neuritis. Neurology, 1998,50: 715-718. Tobimatsu, S., Tomoda, H. and Kato, M. Parvocellular and magnocellular contributions to visual evoked potentials in humans: stimulation with chromatic and achromatic gratings and apparent motion. J. Neurol. Sci., 1995, 134: 73-82. Tobimatsu, S., Shima, F, Ishida, K. and Kato, M. Visual evoked potentials in the vicinity of the optic tract during stereotactic pallidotomy. Electroencephalogr. Clin. Neurophysiol., 1997, 104: 274-279. Tobimatsu, S, Shigeto, H., Arakawa, K. and Kato, M. ElectrophysioJogical studies of parallel visual processing in humans. Electroencephalogr. Clin. Neurophysiol., 1999, 49(Suppl.): 103-107. Tobimatsu, S., Celesia, G.G., Haug, B.A., Onofrj, M., Sartucci, F and Porciatti, V. Recent advances in clinical neurophysiology of vision. Electroencephalogr. Clin. Neurophysiol., 2000, 53(Suppl.): 312-322. Zeki, S.M. A Vision ofthe Brain. Blackwell Scientific Publications, Oxford, 1993. Zeki, S., Watson, lD.G., Lueck, CJ., Friston, KJ., Kennard, C. and Frackowiak, R.SJ. A direct demonstration of functional specialization in human visual cortex. 1. Neurosci., 1991, II: 641--649.