Effect of spatial frequency on simultaneous recorded steady-state pattern electroretinograms and visual evoked potentials

81

Electroencephalography and clinical Neurophysiology, 1991, 80:81-88

© 1991 Elsevier Scientific Publishers Ireland, Ltd. 0168-5597/91/$03.50 ADONIS 016855979100063V EVOPOT 90518

Effect of spatial frequency on simultaneous recorded steady-state pattern electroretinograms and visual evoked potentials Hiroyuki Tomoda, Gastone G. Celesia and Sandra Cone Toleikis Department of Neurology, Loyola University Chicago, Stritch School of Medicine and Hines Veterans Administration Hospital, Maywood, IL 60153 (U.S.A.)

(Accepted for publication: 27 July 1990)

Summary Pattern electroretinograms (P-ERGs) and visual evoked potentials (VEPs) to 4 Hz alternating square-wave gratings were simultaneously recorded in 23 subjects. Responses were Fourier analyzed and amplitude and phase of the 2nd and 4th temporal harmonics were measured. The spatial frequency-amplitude function of the P-ERG 2nd harmonic component displayed either a bandpass tuning behavior, or a low-pass behavior. The peak amplitude for subjects with bandpass tuning was at 1.5 c/deg. The phase of the P-ERG 2nd harmonic decreased monotonically as spatial frequency increased. The VEP 2nd harmonic had a bimodal spatial frequency function with a peak at 3 c/deg and a second increase at spatial frequencies below 1 c/deg, regardless of the P-ERG characteristics. The phase of VEP 2nd and 4th harmonic had an inverted U-shaped function with peak at 3 c/deg and 1.5 c/deg respectively. Comparison of simultaneously recorded P-ERG and VEP spatial frequency functions demonstrated different tuning behavior for cortical and retinal responses. It is concluded that the proposed technique permits the separate analysis of retinal and cortical processing of visual information. The 2nd and 4th harmonic components of VEP behave independently of each other suggesting they may be generated by different subsystems. Key words: Pattern electroretinogram; Visual evoked potentials; Spatial frequency tuning; Phase; Harmonic response

Steady-state evoked potentials are electrophysiological events evoked by rapid sensory stimuli (Regan 1966, 1989; Celesia 1982). Rapid continuous stimulation produces responses of simplified m o r p h o l o g y that are easily quantified in terms of amplitude and phase (Regan 1989). Steady-state EP acquisition is faster and automation facilitated by the application of the Fast Fourier Transform ( F F T ) to the EP recordings (Regan 1966, 1989; Bobak et al. 1983; Porciatti et al. 1989; Celesia and Tobimatsu 1990). Steady-state pattern electroretinograms ( P - E R G s ) and visual evoked potentials (VEPs) have been studied in normals and in various diseases (Spekreijse et al. 1977; Bodis-Wollner et al. 1979; Regan 1983; Baker and Hess 1984; Hess and Baker 1984; Plant and Hess 1986; Plant et al. 1986). Nevertheless, P - E R G s and VEPs are rarely recorded simultaneously to assess the relative contribution of the retina to visually evoked cortical potentials (Sokol et al. 1983; Celesia et al. 1987; Ghilardi et al. 1988).

The aim of the present study was to determine normative values of simultaneously recorded steadystate P - E R G s and VEPs to square-wave gratings and to elucidate their physiological characteristics.

Methods

Twenty-three visually normal subjects, 10 males and 13 females, aged 1 8 - 2 8 (mean age: 24.7) were studied. All subjects were normal volunteers informed of the experimental nature of the test. Every subject had a corrected Snellen visual acuity of 2 0 / 2 0 or better and normal color vision (Ishihara plates), pupil reaction, fundoscopic examination and visual fields. Subjects with refractive error were accepted only if they required lenses with less than 3 diopters correction. Recording

l This study was supported by a grant from the Veterans Administration. Correst~ondence to: Gastone G. Celesia, Department of Neurology, Loyola University Chicago, Stritch School of Medicine, 2160 S. First Avenue, Maywood, IL 60153 (U.S.A.).

P - E R G s were recorded with a bipolar Burian-Allen contact electrode placed over the cornea. The conjunctiva and the cornea were anesthetized with proparacaine hydrochloride 0.5% ophthalmic solution. After electrode placement, each subject was refracted to 2 0 / 2 0 visual acuity. The pupil diameter was measured in each subject.

82

Visual evoked potentials were recorded from silversilver chloride electrodes applied to the scalp with collodion. Electrode impedance was below 3000 12. The occipital electrode was placed on the scalp at the midline, 5 cm above the external occipital protuberance (MO electrode). The MO electrode was referred to a midfrontal electrode 12 cm above the nasion. The ground electrode was at the vertex. Input from corneal and scalp electrodes were fed into preamplifiers adjusted to a bandwidth of 1-250 Hz. Each P-ERG and VEP was the result of the average of 50 responses with a time epoch of 2000 msec. In each subject a "noise run" or "control average" was implemented, with the CRT screen occluded. No discernible response were seen in the control average. F F T of the "control average" failed to detect peaks at the fundamental 2nd and 4th harmonic. The amplitude of the noise at these frequencies was between 0.1 and 0.25/~V. The averaged "noise" response is very similar to the one reported by Porciatti et al. (1989). The signal to noise was estimated (Bobak et al. 1983; Regan 1989) to vary between 1.5 : 1 and 6 : 1 for P-ERG and 2 : 1 and 10 : 1 for VEP.

Stimulation Pattern visual stimulation was carried out in a partially darkened room with an averaged background light of 0.5 c d / m 2. The visual stimuli were vertical gratings with a square-wave luminance profile generated digitally on a Joyce cathode ray tube (CRT) display. The gratings were phase reversed in square-wave fashion at 4 Hz (8 reversals/sec) with a mean luminance of 90 c d / m 2. The contrast between dark and bright bars was 89%. A range of spatial frequencies (0.5-6 c / d e g ) was employed. The stimulating field subtended 7.6 ° .

Signal analysis Signal analysis was performed on-line by a microcomputer. The analog data were digitized at a sampling rate of 256 Hz. Fifty samples of 2 sec epochs were averaged. The averaged response was then analyzed by Fast Fourier Transform (FFT). The F F T resolution was 0.25 Hz. The result of the F F T is a set of values representing the amplitude and phase angle for a particular frequency component. Amplitude and phase were calculated for the 2nd and 4th harmonics. Amplitude was expressed i n / I V and phase in degrees. Phase angle is the angle represented by an arc in a circle, the circle is divided into 360 equal intervals or degrees with no true zero point. The zero point is arbitrary, for example, in a display of the time of the day circularly distributed in equal intervals of 24 h, midnight is set at the zero hour at the top of the circle. Since phase is circularly distributed, the data cannot be evaluated by a linear scale and calculating routine mean and standard deviation would be inappropriate (Mardia 1972; Zar 1974). Statistical methods for analyzing circular distributions were em-

H. TOMODA ET AL.

ployed (Mardia 1972; Zar 1974) and the mean angle, the measure of concentration and the angular dispersion (or circular standard deviation ) were calculated. The mean angle ¢ is expressed by the following equations:

i X=

n cos q~i/n

i=l

Y = Y'~ sin q~i/n i=l

where " n " is number of samples and #'i is phase of each sample r=

X 2 ~

2

where " r " is a "measure of concentration." cos ff = X / r

sin e? = Y / r

¢ = tan-1 Y / X

The range of the mean angle ~ in degrees on a circular scale is estimated by the combination of cos q, and sin q, values. A measure of concentration close to 1 indicates that all the data are in the same direction. A measure of concentration close to zero indicates that there is a great deal of dispersion and that a mean angle cannot be defined. Thus, a measure of concentration close to 1 is desirable and indicative of a very narrow range. The "angular dispersion (circular standard deviation)" is calculated in degrees by the equation: s - 1 8 0 / ¢ r ¢ - 4 . 6 0 5 1 7 log r where log is the common logarithm.

Results Steady-state P-ERG and VEP consisted of quasisinusoidal waves with 2 major F F T peaks: one at the 2nd harmonic component of the stimulus frequency and the other (usually smaller) at the 4th harmonic component (Fig. 1). The characteristics of the spatial frequency functions can be defined in relation to the changes observed along the spatial frequency axis. A function has a low-pass behavior when it shows a decrease limited to high spatial frequencies with no decrease at low spatial frequencies for the parameter under study (i.e., amplitude). Bandpass behavior or tuning indicates a decrease at both the lower and the upper end of the spectrum of spatial frequencies studied. Bandpass behavior or tuning will usually show a preferred spatial frequency or a range of preferred spatial frequencies represented as a peak in the function.

(1) P-ERG The P-ERG 2nd harmonic responses were of small amplitude (Table I). The spatial frequency functions demonstrated either bandpass or low-pass behavior (Fig. 2). In 14 subjects, the amplitude showed a monotonic decrease with increasing spatial frequency, with a more marked decrease in amplitude at frequencies above 3 c / d e g (low-pass behavior). The remaining 9 subjects

STEADY-STATE P-ERGs A N D VEPs

83

cpd

1.5

ERG 2F

VEP

VEP

2F 4F

4F

6.0

ERG

"~ 4.0 --'~ 3.0

cpd

3.0

qJ z

w 2.0 O' tll

~,, 1.5

VEP

i1 .,J

_~ 1.0

ERG

a.

.5

i 2~V I

I

I

I

200

I

I

I

MSEC/DIV

I

I

~ I

I

~ 1

I

v l

I

I

I

I

k l l l l l l l l

I

FREQUENCY(6.1 Hz/div)

Fig. 1. The column on the left shows pattern VEP and ERG for stimuli at 1.5 and 3.0 c / d e g (cpd). Two reduplications of the averaged signals are shown to demonstrate reproducibility of the responses. On the fight are depicted the amplitude spectra of the simultaneously recorded P-ERG and VEP at various spatial frequencies. Note that the 4th harmonic (4F) of the VEP behaves differently than the 2nd harmonic (2F).

showed a bandpass spatial frequency-amplitude function with a preferred peak at 1.5 c/deg. The 4th harmonic P-ERG component behaves essentially in the

same fashion as the 2nd harmonic component, showing a low-pass behavior with a decrease at frequencies above 3 c / d e g or a bandpass function with a preferred peak at

TABLE I Amplitude in /tV and phase in degree of P-ERG and VEP. S.E. = standard error; AD = angular dispersion; r = measure of concentration. Spatial frequency (c/deg)

Amplitude (/~V)

Phase (°)

Second harmonic

Fourth harmonic

Ratio ( 2 F / 4 F )

Second harmonic

Mean

S.E.

Mean

S.E.

Mean

S.E.

Mean

AD

r

P-ERG 0.5 0.7 1.0 1.5 2.0 3.0 4.0 6.0

0.29 0.31 0.29 0.32 0.26 0.26 0.25 0.16

0.02 0.03 0.02 0.02 0.02 0.02 0.02 0.02

0.16 0.13 0.13 0.15 0.13 0.13 0.12 0.08

0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.01

2.11 2.47 2.79 2.32 2.45 2.41 2.31 2.30

0.17 0.13 0.33 0.14 0.23 0.19 0.14 0.21

- 14.7 -20.3 - 17.6 - 21.1 - 26,5 - 33.8 - 25.9 - 29.6

16.9 20.7 20.6 22.3 27.0 25.4 23.0 34.5

0.5 0.7 1.0 1.5 2.0 3.0 4.0 6.0

1.22 1.17 1.10 1.11 1.14 1.31 1.23 0.85

0.09 0.10 0.08 0.12 0.09 0.11 0.12 0.08

0.36 0.35 0.42 0.64 0.72 0.83 0.73 0.55

0.05 0.05 0.06 0.09 0.08 0.09 0.08 0.07

5.15 4.74 4.41 2.43 2.15 1.99 2.12 1.88

0.81 0.68 1.05 0.38 0.30 0.26 0.25 0.19

148.3 154.2 157.5 172.2 173.4 174.6 168.9 159.1

21.7 18.2 20.3 30.2 28.9 25.9 26.8 30.2

VEP

Fourth harmonic Mean

AD

r

0.96 0.94 0.94 0.93 0.90 0.91 0.92 0.83

88.0 82.4 73.0 59.9 57.1 53.9 46.2 36.4

14.4 14.5 12.7 21.6 14.4 19.9 12.1 25.8

0.97 0.97 0.98 0.93 0.97 0.94 0.98 0.90

0.93 0.95 0.94 0.87 0.88 0.90 0.90 0.87

117.3 147.0 169.8 178.0 171.9 141.8 117.6 80.6

85.2 82.1 66.1 39.5 41.2 37.6 44.9 55.4

0.33 0.36 0.51 0.79 0,77 0.81 0.74 0.63

84

H. T O M O D A ET AL.

1.5 c/deg. The amplitude ratio of the P-ERG 2nd and 4th harmonic response, calculated to determine the relative proportion of the 2nd and 4th harmonic response, remained relatively constant over the full range of spatial frequencies (Table I). In the search for a variable that may explain the different behavior of the P-ERG spatial frequency functions between the two groups the pupillary diameter was analyzed. The mean and standard deviation of the pupillary diameter for the group with bandpass functions was 4.4 _+ 0.44 mm, whereas for the group with low-pass functions was 5 . 1 _ 0.33 ram. The difference between the two groups was statistically significant (t test, P < 0.05). As shown in Fig. 3, the phase of the P-ERG 2nd and 4th harmonic response decreased monotonically as spatial frequency increased, regardless of the bandpass or low-pass characteristic of the P-ERG 2nd harmonic amplitude. The phase of the 2nd harmonic (see Table I) varied in different individuals but only within a very narrow range. At each spatial frequency, the measure of concentration and the angular dispersion for the group indicated that the data were in the same direction and varied within a desirable narrow range. As shown in ERG

8ANOPASS

.41

. . . . . . . .

i

. . . . . . . .

l

~ .3

.4

i

. . . . . . . .

i

.

L

.3 .2

4F

0 O.

.I

i ,

......

i

. . . . . . . .

1

i

i

i.

....

O.

10

. .......

10

1

~

.

.

.

.

.

.

.

.

.

.

rl"

I'B 1

2F

>~:~

1.1 .II

~1.1

2F

UA Q

.61

.6

! .1

,

0.1

i

......

i

1 C/O

.

. . . . . . . .

10

350 340

330

N

VEP 2F IgC "

. . . . . . . .

%

. . . . . .

i

......

170 180

i

. . . .

±..4J

10

1

140 O.

ERG 4F

. . . . . . . .

i

80

170 k

i

/

eo

i+o~

40

IIOF

20

so 50 h ........ 0. 1

, 1

i

. . . . . . . .

10

VEP 4 F ........ '

......

. . . . . . . .

1

200 H

I00 .r

. . . . . . . .

150

O.

"~

•

IBO

320 310

'

........ t ~

L~

CL

0 0. 1

...... , . . . . . . . . l 10 CIO

........

l0 CID Fig. 3. Spatial frequency functions of the P-ERG and VEP phase of all subjects. The upper graphs represent the mean and standard error of phase of the 2nd harmonic (2F), the lower graphs the 4th harmonic (4F). The left column refers to P-ERG and the fight column to VEP. Note the monotonic decrease of the P-ERG functions, whereas the VEP functions show an inverted U-shaped curve with a peak at 3 c / d e g for 2nd harmonic and 1.5 c / d e g for 4th harmonic.

Fig. 4 the boundary of normality could be established for the phase at the 95% confidence limit. The phase of the P-ERG appears 180 ° opposite to the VEP (Fig. 4), raising the question whether we were recording volume conducted VEP. In 2 subjects with acute optic neuritis, we recorded normal P-ERG, whereas VEP was absent, thus confirming that the P-ERG signal originated in the retina.

(2) VEP

VEP l°

360

i

LOW-PASS . . . . . . . .

w

4F

ERG 2F . . . . . . . .

l

,F . . . . . . . .

O.

I

1 C/O

. . . . . . . .

IO

Fig. 2. Spatial frequency functions of the P-ERG and VEP amplitude. The two upper graphs represent the mean amplitude and the standard error of the P-ERG across subjects, the lower graphs represent the VEP. The left column refers to subjects with P-ERG bandpass tuning behavior, the right column refers to subjects with P-ERG bandpass behavior. Note that the VEP 2nd harmonic (2F} shows bimodal distribution, regardless of the P-ERG characteristics. The P-ERG 4th harmonic (4F) behaves essentially in the same fashion as the 2nd harmonic, whereas the VEP 4th harmonic behaves differently from the 2nd harmonic.

Unlike the P-ERG, the amplitude and spatial frequency function of the 2nd harmonic was bimodal with a peak at 3 c / d e g (Fig. 2), and a second increase at spatial frequencies below 1 c/deg. The mean amplitude of the 4th harmonic showed a classic bandpass behavior with a preferred tuning at 3 c/deg. This bimodal distribution occurred in each subject regardless of the bandpass or low-pass function of the simultaneously recorded P-ERG. The amplitude ratio of the 2nd and 4th harmonic VEP responses showed a sharp decrease at spatial frequencies above 1 c / d e g (Table I), suggesting that the 4th harmonic response behaved differently from the 2nd harmonic as the spatial frequency increased. The phase of the 2nd harmonic showed an inverted U-shaped behavior with an increase up to 3 c / d e g followed by a decrease at higher spatial frequencies. The phase of the 4th harmonic showed a similar inverted U-shaped behavior with, however, a peak at a

STEADY-STATE P-ERGs AND VEPs

85

ERG 2F .5 C/D

.7 C/D

1.0 C/D

1.5 C/D

0

0

0

0

180

180

2 o 9o 2 o 9o

180

180

0.6

2.0 C/D

3.0 C/D

4.0 C/D

6.0 C/D

0

0

0

0

0.0

270

90 270

180

90 270

180

90

270

180

90

180

VEP 2F .5 C/D

.7 C/D

1.0 C/D

1.5 C/D

27o 9o27o 9020 9o 20 9o 0

0

0

180

180

2.0]

2.0 C/D

3.0 C/D

o.o

27o 9o 7o 90 0

0

0

180 I laV

180

4.0 C/D 0

6.0 C/D

90

0

90

180 180 180 180 Fig. 4. Polar graphs of the P-ERG and VEP 2nd harmonic amplitude and phase in all subjects. The radial axis indicates the amplitude in microvolts, the circular axis represents the phase in degrees. The data at each spatial frequencies are represented by different circles. The data of each subject are represented by open circles. The boundary of normality of the phase was established at the 95% confidence limit. Note that the phase of the 2nd harmonic showed a narrow range of values at each spatial frequency, confirming the reliability of the phase as suggested by measure of concentration "r" in Table I.

86

H. TOMODA ET AL.

ERG 2F

VEP 2F

•

/

;j

ERG 4F

o

180

2"°1

2

VEP 4F

o

180

Fig. 5. Polar graphs of amplitude and phase in one subject. The radial axis indicates the amplitude in microvolts, the circular axis represents the phase in degrees. The data at each spatial frequency are represented by filled circles and the arrows indicate the direction of the function from 0.5 c / d e g to 6 c/deg. Note the bandpass characteristic of the 2nd (2F) and 4th (4F) harmonic of the VEP, and the low-pass characteristic of the 4th harmonic of the P-ERG, whereas the 2nd harmonic response of the P-ERG has a weak bandpass characteristic.

lower frequency of 1.5 c/deg (Fig. 3). The amplitude and phase functions of the VEP 2nd and 4th harmonic components showed the same characteristics regardless of the bandpass or low-pass function of the simultaneously recorded P-ERG. The relationship of the 2nd and 4th harmonic amplitude and phase for one subject is illustrated in Fig. 5. The polar graphs show the strong tuning for the 2nd and 4th harmonic components of the VEP. There is a response over a wide variety of spatial frequencies but the response is twice as large for the preferred frequency than for the non-preferred spatial frequency. The tuning in this subject showed a peak at 4 c/deg for the 2nd harmonic and at 2 c/deg for the 4th harmonic. The phase of the 2nd harmonic for the group showed a narrow range of values at each spatial frequency (Fig. 4). The measure of concentration and the angular dispersion data confirmed the reliability of the phase (Table I). The boundary of normality were established (Fig. 4) at the 95% confidence limit.

Discussion

The present study has shown the feasibility to simultaneously record steady-state P-ERG and VEP in humans. Both signals could be characterized by their

amplitude and phase. Boundaries of normality could be established at the 95% confidence limit for amplitude and phase. Phase showed a narrow angular dispersion. Therefore it cab be utilized as a measurement similar to latenlcy for transient P-ERG and VEP. Regan (1989) warned that latency in steady-state evoked potentials cannot be estimated unequivocally because there are non-linearities in the ~system under study that may contribute to the value of the phase. Pragmatically, however, the precise measurement of the phase under the same recording and stimulation paradigms can be utilized in the same fashion as the peak latency is used for transient responses (Riemslag et al. 1982; Tobimatsu et al. 1990). FFT analysis of steady-state responses can be utilized in clinical setting to objectively establish the spatial frequency functions of humans. Abnormalities can be defined either as phase shifts or amplitude depressions at a given spatial frequency. Neima and Regan (1984) have shown psychophysical impairment at selective frequencies in patients with multiple sclerosis. Our group (Celesia and Brigell 1990) has similarly demonstrated impairment of transient VEP at selective spatial frequencies in optic neuritis patients. Spatial frequency functions evaluated with transient VEPs, how#er, were found to be impractical because of the length of the testing. The relative speed of data acquisition using FFT analysis of steady-state VEP makes the method pragmatically feasible. The present data show that P-ERG behaved differently in different subjects. Spatial bandpass tuning was observed in one group of subjects, while low-pass functions were observed in a second group. This discrepancy may explain the controversy in the literature between the authors reporting spatial tuning of P-ERG amplitude (Fiorentini et al. 1981; Hess and Baker 1984; Plant and Hess 1986; Plant et al. 1986; Porciatti et al. 1988) and others reporting low-pass characteristics (Armington and Brigell 1981; Riemslag et al. 1985). The reason for the two different spatial frequency functions remains to be explained. Odom et al. (1982) noted that experiments failing to document spatial tuning have used higher luminance level. These authors suggested that at high luminance there is masking of the contrast response (responsible for the tuning) by the local luminance response. Holopigian et al. (1988) attributed these conflicting results to the large amount of variability inherent in amplitude measurements. Both factors probably played a role in our data. Changes in pupillary size modify, at times drastically, retinal illuminance and therefore the response characteristics of P-ERG (Tobimatsu et al. 1988). In the present study, subjects with P-ERG low-pass behavior had a significantly larger pupil size when compared to those with P-ERG bandpass behavior. It is possible that a larger pupil size and its related increased retinal illuminance may have

STEADY-STATE P-ERGs AND VEPs

masked the spatial frequency tuning in some of our subjects. Increased retinal illuminance will result in greater light scatter and thus increased luminance contamination of P-ERG responses (Hess and Baker 1984). The possibility that spatial frequency tuning was masked by the luminance response is suggested by the presence of a small peak in the amplitude response at 1.5 c/deg in the group with low-pass functions. We conclude that P-ERG to square-wave gratings is a mixed response containing both luminance and contrast responses. The 4th harmonic P-ERG component behaves essentially in the same fashion as the 2nd harmonic component. The monotonic decrease of phase as spatial frequency increases confirms the data of other authors (Plant and Hess 1986; Plant et al. 1986; Porciatti et al. 1988). The decrease in phase is probably related to the optic degradation of higher spatial frequency patterns (Campbell and Green 1965; Korth and Rix 1988). Amplitude and phase functions of steady-state VEP were different than the simultaneously recorded P-ERG. The amplitude of the 2nd harmonic component of the VEP showed a bimodal distribution with a peak at 3 c/deg and a second increase at lower spatial frequencies, regardless of the bandpass or low-pass characteristics of the P-ERG 2nd harmonic function. In the group with P-ERG spatial bandpass tuning, the preferred peak occurred at 1.5 c/deg at the retinal level, whereas, it occurred at 3 c/deg at the cortical level. The preferred response at 3 c/deg is consistent with previous studies (Ristanovic and Hajdukovic 1981; Regan 1983; Hess and Baker 1984; Plant and Hess 1986; Plant et al. 1986). The phase data with their U-shaped function and the amplitude ratio 2nd/4th harmonic suggest that 2 mechanisms are operating: one at low frequency and another at mid and high frequencies. The bimodal distribution of the response as a function of spatial frequency may represent a mixture of pattern-contrast and luminance responses (Odom et al. 1982). At low frequencies below 1.5 c/deg, the amplitude increases as the frequency decreases thus suggesting a luminance response. Responses to medium and high spatial frequencies have a bandpass function suggesting a mechanism related to pattern-contrast. The amplitude of the 4th harmonic component showed a classic bandpass tuning. The present data on the spatial tuning function of the 2nd and 4th harmonic components demonstrate that the 4th harmonic behaves differently of the 2nd harmonic component. The 4th harmonic component is proportionally larger when compared to the 2nd harmonic at higher spatial frequencies. This might suggest that the 4th harmonic VEP component represents, at least partially, a distinct physiological response. The bandpass characteristic of the 4th harmonic response further suggests that this response is less contaminated by luminance activation. Whether the 4th harmonic response is generated by

87

separate or overlapping neuronal pools than those generating the 2nd harmonic response remains to be clarified. Comparison of the simultaneously recorded P-ERG and VEP spatial frequency functions demonstrated a different tuning behavior for the cortical responses than the one generated in the retina. It is concluded that the proposed technique permits the separate analysis of retinal and cortical processing of visual information. The analysis of the 2nd and 4th harmonic components of these potentials further permit the study of responses generated by different physiological subsystems.

References Armington, J.C. and Brigell, M. Effects of stimulus location and pattern upon visually evoked cortical potentials and the electroretinogram. Int. J. Neurosci., 1981, 13: 169-178. Baker, C.L. and Hess, R.F. Linear and non-linear components of the human electroretinogram. J. Neurophysiol., 1984, 51: 952-967. Bobak, P., Bodis-Wollner, I., Harnois, C., Maffei, L., Mylin, L., Podos, S. and Thornton, J. Pattern electroretinograms and visually-evoked potentials in glaucoma and multiple sclerosis. Am. J. Ophthalmol., 1983, 96: 72-83. Bodis-Wollner, I., Hendley, C.D., Mylin, L.H. and Thornton, J. Visual evoked potentials and the visuogram in multiple sclerosis. Ann. Neurol., 1979, 5: 40-47. Campbell, F.W. and Green, D.G. Optical and retinal factors affecting visual resolution. J. Physiol. (Lond.), 1965, 181: 576-593. Celesia, G.G. Steady state and transient visual evoked potentials in clinical practice. Ann. NY Acad. Sci., 1982, 388: 648-650. Celesia, G.G. and Brigell, M.G. Pattern visual stimulation in prechiasmatic lesions. In: J.E. Desmedt (Ed.), Visual Evoked Potentials. Clinical Neurophysiology Updates, Vol. 3. Elsevier, Amsterdam, 1990: 75-85. Celesia, G.G. and Tobimatsu, S. Electroretinograms to flash and to patterned visual stimuli in retinal and optic nerve disorders. In: J.E. Desmedt (Ed.), Visual Evoked Potentials. Clinical Neurophysiology Updates. Elsevier, Amsterdam, 1990: 45-55. Celesia, G.G., Kaufman, D. and Cone, S. Effects of age and sex on pattern electroretinograms and visual evoked potentials. Electroenceph, clin. Neurophysiol., 1987, 68: 161-171. Fiorentini, A.L., Maffei, M., Pirchio, V. and Spinelli, D. The ERG in response to alternating grating in patients with diseases of the peripheral visual pathway. Invest. Ophthalmol. Vis. Sci., 1981, 21: 490-493. Ghilardi, M.F., Bodis-Wollner, I., Onofrj, M.C., Marx, M.S. and Glover, A.A. Spatial frequency-dependent abnormalities of the pattern electroretinogram and visual evoked potentials in a parkinsonian monkey model. Brain, 1988, 111: 131-149. Hess, R.F. and Baker, Jr., C.L. Human pattern-evoked electroretinogram. J. Neurophysiol., 1984, 51: 939-951. Holopigian, K., Snow, J., Seiple, W. and Siegle, I. Variability of the pattern electroretinogram. Docum. Ophthalmol., 1988, 70: 103115. Korth, M. and Rix, R. Luminance-contrast evoked responses and color-contrast evoked responses in the human electroretinogram. Vision Res., 1988, 28: 41-48. Mardia, K.V. Statistics of Directional Data. Academic Press, London, 1972. Neima, D. and Regan, D. Pattern visual evoked potentials and spatial vision in retrobulbar neuritis and multiple sclerosis. Arch. Neurol., 1984, 41: 198-201.

88 Odom, J.V., Madia, T.M. and Dawson, W.W. Pattern evoked retinal response (PERR) in human: effect of spatial frequency, temporal frequency, luminance and defocus. Curr. Eye Res., 1982, 2: 99-108. Plant, G.T. and Hess, R.F. The electrophysiological assessment of optic neuritis. In: R.F. Hess and G.T. Plant (Eds.), Optic Neuritis. Cambridge University Press, Cambridge, 1986: 192-229. Plant, G.T., Hess, R.F. and Thomas, S.J. The pattern evoked electroretinogram in optic neuritis. A combined psychoph~'sical and electrophysiological study. Brain, 1986, 109: 469-490. Porciatti, V., Falsini, B., Scalia, G., Fadda, A. and Fontanesi, G. The pattern electroretinogram by skin electrodes: effect of spatial frequency and age. Docum. Ophthalmol., 1988, 70: 117-122. Porciatti, V., Falsini, B., Fadda, A. and Bolzani, R. Steady-state analysis of the focal ERG to pattern and flicker: relationship between ERG components and retinal pathology. Clin. Vis. Sci., 1989, 4: 323-332. Regan, D. Some characteristics of average steady-state and transient responses evoked by modulated light. Electroenceph. clin. Neurophysiol., 1966, 20: 238-248. Regan, D. Spatial frequency mechanisms in human vision: electrophysiological evidence. Vision Res., 1983, 23: 1401-1407. Regan, D. Human Brain Electrophysiology, Elsevier, New York, 1989: 1-8, 34-43, 302-436. Riemslag, F.C.C., Spekreijse, H. and Van Walbeek, H. Pattern evoked potential diagnosis of multiple sclerosis: a comparison of various contrast stimuli. In: J. Courjon, F. Maugui6re and M. Revol

H. TOMODA ET AL. (Eds.), Clinical Applications of Evoked Potentials in Neurology. Raven Press, New York, 1982: 417-426. Riemslag, F.C.C., Ringo, J.L., Spekreijse, H. and Verduyn Lunel, H.F. The luminance origin of the pattern electroretinogram in man. J. Physiol. (Lond.), 1985, 363: 191-209. Ristanovic, D. and Hajdukovic, R. Effects of spatially structured stimulus fields on pattern reversal visual evoked potentials. Electroenceph, clin. Neurophysiol., 1981, 51: 599-610. Sokol, S., Jones, K. and Nadler, D. Comparison of the spatial response properties of the human retina and cortex as measured by simultaneously recorded pattern ERGs and VEPs. Vision Res., 1983, 23: 723-727. Spekreijse, H., Estevez, O. and Reits, D. Visual evoked potentials and the physiological analysis of visual processes in man. In: J.E. Desmedt (Ed.), Visual Evoked Potentials in Man. Clarendon, Oxford, 1977: 16-89. Tobimatsu, S., Celesia, G.G. and Cone, S.B. Effects of pupil diameter and luminance changes on pattern electroretinograms and visual evoked potentials. Clin. Vis. Sci., 1988, 2: 293-302. Tobimatsu, S., Tashima, S., Hiromatsu, M. and Kato, M. Clinical feasibility of steady-state VEPs: relationship between phase and the P100 latency of transient VEPs. Electroenceph. clin. Neurophysiol., 1990, 75: S151. Zar, J.H. Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, N J, 1974: 310-328.