0042-6989/91$3.09+ 0.00 Copyright 0 1991 Pergsmon F’ress plc

Won Res. Vol. 31, No. 10, pp.18314836,1991 Printed in Great Britain. All rights reserved

RESEARCH NOTE

VISUAL FIELD ASYMMETRIES IN PATTERN DISCRIMINATION: A SIGN OF ASYMMETRY IN CORTICAL VISUAL FIELD REPRESENTATION? NICOLETTA BERABDI and ADR~ANAFIORENTINI Istituto di Neurofisiologia CNR, Via S. Zeno 51, Pisa, Italy (Received 8 October 1990; in revised form

4 February 1991)

Abatraet-A visual field asymmetry is described relative to the discrimination of mirror symmetric bars with ramp-like luminance profiles. Along the vertical meridian the discrimination is better performed for patterns oriented parallel to the meridian than for patterns oriented orthogonally at all eccentricities tested (2-g deg). Along the horizontal meridian, the preference for radially oriented stimuli is present at 2 deg from the fovea, but vanishes at larger eccentricities. The meridional asymmetry thus revealed psychophysically may reflect asymmetries in the representation of the vertical and horizontal meridians in the human visual cortex. Peripheral vision

Pattern discrimination

Spatial phase

INTRODUCTION

Orientational anisotropy

profiles (see insert in Fig. 1). Discrimination of complex gratings in peripheral vision has been widely investigated (Braddick, 1981; Rentschler & Treutwein, 1985; Bennett dc Banks, 1987; Berardi & Fiorentini, 1987; Morrone, Burr & Spinelli, 1989) and there are indications that it can be locally anisotropic, i.e. the discrimination performance can be different according to pattern orientation (Berardi & Fiorentini, 1988). Orientational biases have been reported also for the discrimination of aperiodical stimuli having mirror-symmetric luminance profiles (Bennett & Banks, 1991). It will be shown that, for the present discrimination task, the orientational biases are radially symmetric in the parafovea (2 deg eccentricity), while at larger eccentricities they are preserved along the vertical, but not along the horizontal meridian. nance

It is well known that visual performance varies considerably from the centre towards the periphery of the visual field. For most resolution tasks there is no gross asymmetry in the central portion of the visual field: within 5-10 deg from the fovea the decline of performance with eccentricity is approximately the same along the various meridians (Rovamo & Virsu, 1979; Pointer 8z Hess, 1989). For pattern discrimination tasks, asymmetries can be observed even at small distances from the fovea, such as a right or left hemifield superiority, related to left or right hemispheric dominance (Davidoff, 1982). For instance, a left visual field superiority has been reported in the discrimination of complex gratings differing for the relative phase of their harmonic components (Fiorentini & Berardi, 1984). We report here a pattern discrimination task that reveals a marked asymmetry in performance between the vertical and the horizontal meridians. The patterns to be discriminated were aperiodical and were obtained from complex gratings (fundamental plus second harmonic with relative spatial phases 90 or 270 deg) screened by a square window with side corresponding to 1.2 periods of the fundamental component. The resulting patterns had mirror-symmetric lumi-

METHODS

The patterns were computer generated (PDPl l/03) on a Tektronix 608 display and were surrounded by a uniform field of the same mean luminance (40 cd/m’). They consisted in 1.2 periods of gratings having luminance profiles:

1831

L = L,[l + m cos(2 nfx) + l/2 m cos(2 x 2 fx + d)]

1832

Research Note

with 4 = 90 and 270 deg, respectively. The period of the fundamental component was 20 mm and the stimuli were limited by a square window 24mm side. The truncated luminance profiles were approximately mirror symmetric. The stimuli could be located at three different distances from the fixation point (2,4 and 8 deg, measured from the inner edge of the window) either along the horizontal or the vertical meridian of the visual field and could be oriented parallel or orthogonal to the meridian. The size of the stimuli was scaled according to the presumed human magnification factor (Tolhurst & Ling, 1988) to compensate for changes in detectability of the stimuli with eccentricity (viewing distances: 180,90 and 60 cm for observations at 2, 4 and 8 deg eccentricity, respectiveIy). Viewing was binocular and refraction corrected, if necessary, for the viewing distance. Stimulus presentation was under computer control. First, contrast detection thresholds were measured for each observer, with a temporal forced-choice (pattern vs uniform field) multiple staircase procedure converging at 82% correct. Stimulus duration was 1OOmsec. Detection thresholds for vertical and horizontal stimuli were compared at each eccentricity and were found never to differ more than 0.1 log units. Then discrimination of mirror symmetric patterns was tested for each eccentricity, visual field location and stimulus orientation, by determining the percentage of forced-choice correct responses as a function of stimulus peak-topeak contrast. 4-8 blocks of 40 trials each were performed for various contrast values, exceeding the detection threshold by different amounts. In each trial two patterns were presented for 100 msec each, with an interstimulus interval of 1 sec. The two patterns could have the same or d&rent (symmetric) luminance proflle. The subject’s task was to signal “same” or “different” by pushing one of two buttons. Pushing a third button started a new trial. For each condition the initial 100 trials were discarded to allow for training effects (Fiorentini & Berardi, 1980, 1981). The psychometric curves obtained as a function of stimulus contrast were linearized by transforming the percent correct into Normal Equivalent Deviates (Finney, 1971). RESULTS AND DlSCUSSl~N Three subjects took part in the experiment. Two (AF and NB) were experienced subjects,

well used to discrimination in the peripheral visual field, the third (TP) was a naive subject whose fixation was checked during trials. At 2deg eccentricity patterns oriented radially, i.e. parallel to the visual field meridian, were better discriminated than patterns oriented orthogonally to the meridian. This is shown for subject AF in Fig. 1, where psychometric functions are reported for stimuli of vertical or horizontal orientation at 2 deg eccentricity in the left (A), right (B), lower (C) and upper (D) hemifields. For each field location the psychometric functions for the two orientations are clearly different. The percent correct for stimuli parallel to the meridian is significantly larger than for orthogonal stimuli for any contrast value yielding performances better than chance. In addition, the computed regression lines (A, B and C) for the nonpreferred orientation are shallower. Taking the 75% correct point as the contrast threshold for discrimination, in three hemifields (left, right, lower) the threshold is about a factor of 2 higher for stimuli oriented orthogonally to the meridian than for radial stimuli. The difference is even larger for the upper visual field, where the horizontal patterns could hardly be discriminated even at a very high contrast. The data shown in Fig. 1, A and B, confirm the previously observed asymmetry between the left and the right visual hemified, probably reflecting hemispheric specialization (Fiorentini & Berardi, 1984). A superiority of the lower hemifield with respect to the upper hemifield was also observed in the present discrimination task. Similar results were obtained in the other two subjects. Examples of psychometric functions of subject TP at 2 deg eccentricity are presented in Fig. 2, A and B. It has to be noted that in fovea1 viewing vertical and horizontal complex grating differing for their luminance profiles are discriminated equally well (Fiorentini 8z Berardi, 1981). We extended this finding to the discrimination of the aperiodical stimuli employed in the present work by measuring the percentage of correct responses for fovea1 horizontal and vertical patterns of equal contrast. No significant orientational bias was obtained in fovea1 discriminations. The preference for radially oriented patterns at 2deg eccentricity was found to hold in all subjects not only for the vertical and horizontal meridian but also for an oblique meridian.

f833

Research Note 2’

r

(Af

r

Upper field

Lower field

Right field

Left field 99.6

Eccentricity

r

f (Cl

1111

10)

_

0 Velticol 0 ~orizantol

95 90 80

--i-/

TO 0

60

I

I

I

0.6

0 8

1.0

0.4

0.6

0.8

1.0

,

,

*

l/

f

J

0.6

0.8

1.0

8 * Eccentricity Left field

Right field

Lower field

Upper firdd

SC A

95 90 80 70 60

Contrast &NJ units above detection thr@&Wtd)

Fig. I. Psychometric functions for the temporal forced choice discrimination of two aperiodical patterns with mirror symmetric luminance profiles (see insert). The pattern eccentricity was 2 deg (top) and 8 deg (bottom). The percentage of correct responses is reported as a function of the peak-to-peak contrast of the stimuli in log units above the, corresponding detection threshold. Number of trials per point: 160 (A, C and H); 200 (B and G, open symbols); 240 (D); 320 (E and F); 280 (G, solid symbols). The sohd circles in H indicate that performance was at chance level (50% correct). Tests of significance were performed for psychometric functions with at least three data points. The thresholds for the preferred and nonpreferred orientation differ significantly in A, B, C and G (two-tailed t-test, P < 0.01). The slopes of the corresponding regression lines also d&r s~~i~~~t~y (two-tailed t-test, P < O.Qt). No si~ifi~nt difference is found in E and F.

In the lower left visual field quadrant the dise~minatjon threshold was smaller for patterns oriented at 45 deg from vertical and pointing toward the fovea (radial orientation) than for orthogonal patterns (tangential orientation). We conclude that at 2deg eccentricity there is an orientational anisotropy that obeys the radial symmetry rule: stimuli pointing to the fovea are better disseminate than orthogonal stimuli for any meridian tested, A possible explanation for radial/tangential anisotropies is that the fixation point is a better spatial referent in the radial than in the tangential viewing condition. If this were the case for our discrimination task, one should expect the preference for radial patterns to be preserved at larger eccentricities, That this is not true is shown for subject AF in Fig. 1, E-H, where the psychometric

functions are reported for vertical and horizontal stimuli at 8 deg eccentricity% For stimuli located along the vertical meridian (Fig. 1 G and H) the results at 8 deg replicate the preference for the vertical orientation observed at smaller eccentricities. In particular, in the upper hemifield the performance was significantly better than chance for vertical stimuli, while for horizontal stimuli it was still at chance level at the highest available contrast (l-4 log units above detection threshold). Along the horizontal meridian on the contrary (Fig. 1, E and F), the preference for the horizontal orientation is not preserved at 8 deg. Indeed the psychometric functions for vertical and horizontal stimuli are superimposed one to each other, both in the left and in the right hemifield. The findings of AF at 8 deg eccentricity were replicated by the other two subjects

1834

Research Note Left

field

Lower 2’

66

field

Left

-

Lower 4.

-

(A)

field

EccH?tlicity

Sbj TP

(8)

r

. 95 -

field

Eccentricity

Sbl

(E)

NE

0

/’ /

SO/f

60-1

0.6

f

0.0

1.0

1.2

0.4

0.6

0.6

‘1.0

0.4

0.6

0.8

1.0

0.6

0.8

1.0

1.2

0 6. Eccentricity

CF Eccwltricity

Z 69(C)

r

0

(0)

95 so -

/

60-

l

/’

, 1.0

,

,

1.2

1.4

-

, 1.0 Contrort

f

I

1.2

1.4

(log

unite

1.0 obovo

detection

1.2

1.4

1.0

1.2

1.4

thmhold)

Fig. 2. Examples of psychometric functions for subjects TP (A-D) and NB (E-H) for the dkimination of horizontal (a) or vertical (0) mirrorqmmetrk patterns located in w kft (A. C, E, G) or lowr (B, D,F,H)vi~fteldatecoentricitiaof:2des(AradB);4dep@andF);8del(C,D,G,ki).Numkr of trials per point: I20 (A-D, G and H); t&J (E and F).

(Fig. 2, bottom row). At an intermediate eccentricity, (4&g) an asymmetry between the vertical and horizontal meridian is already present (see Fig. 2, E and F) albeit less marked than at 8 deg. Figure 3 summa&es the results for the three subjects. The dWrenee in contrast threshold for the discrimination of radial versus tangential stimuli is plotted against eccentricity for the horizontal (solid symbols) and vertical (open symbols) half meridian. The contrasting effects of increasing eccentricity along the two meridians is evident from the figure: the orientational anisotropy in stimulus discrimination is constant from 2 up to 8 deg along the vertical meridian while it progressively decreases along the horizontal meridian to vanish at 8deg eccentricity. Orientational anisotropies have been described for other visual tasks, such as visual acuity (Rovamo, Virsu, Laurinen & Hyvarinen, 1982), contrast sensitivity (Pointer UC Hess, 1989), curvature detection (Fahle, 19%) and motion detection (McColgin, 1960). These

ECCENTRICITY

(dog)

for the Fig. 3. Difkencc bWvwu the log ~~~tmst dis&mimitionofstimulio&ntaipnibltothemsrldhn and the log thfesh~ki for sthdi p!pdidU TV Ihe meridian,@ttedasafu8Uionof~ty.~~bols kdicate stimali pacated aiong tk iowsr wtkal lnericw,soiiipgmbokstirmdiprraatbd~~~ horizontal maridkn. SubjecU: NB (cirdea); AF (srW=s); TP (triangles).

1835

ResearchNote

Bennett, P. J. & Banks, M. S. (1991). The efkc~s of anisotropies become significant only beyond contrast, spatial scale and orientation on fovea1 15-20 deg from the fovea, with one notable and peripheral phase discrimination. Vision Research, exception for a hyperacuity task (Yap, Levi 31, 175!%1786. & Klein, 1987). In all cases, performance is Berardi, N. & Fiorentini, F. (1987). Interhemispheric transfer of visual information in humans: Spatial better for radially oriented stimuli along characteristics. Journal of Physiology, London, 384, any meridian. The preference for radial 633-647. stimuli may result from local anisotropic Berardi, N. & Fiorentini, A. (1988). Lateralization and properties of the human visual cortex. In orientational bias in the discrimination of mirror-symmetthe monkey parafoveal striate cortex, an orienric complex gratings. Perception, 17, 364. tational bias has been reported: cells of Braddick, 0. (1981). Is spatial phase degraded in peripheral vision and visual pathology? In Maffei, L. (Ed.), Pathothe supragranular layers tuned to radial or physiology of the visual system (pp. 255-262). Documenta nearly radial orientations are present in a higher Ophthalmologica Proceedings Series, no. 30. The Hague: proportion than cells tuned to orthogonal W. Junk. orientations (Bauer & Dow, 1989). This in turn Davidoff, J. (1982). Studies with non-verbal stimuli. In Beaumont, J. G. (Ed.), Divided visual field studies of may reflect an orientational bias of retinal cerebral organization (pp. 29-55). London: Academic receptive fields (Levick & Thibos, 1982; Press. Leventhal & Schall, 1983; Schall, Perry & Fahle, M. (1986). Curvature detection in the visual field and Leventhal, 1986). a possible physiological correlate Experimental Brain The anisotropic properties revealed by the Research, 63, 113-I 24. present discrimination task, however, are radi- Finney, D. J. (1971). Probit analysis. Cambridge: Cambridge University Press. ally symmetrical in the visual field only within Fiorentini, A. & Berardi, N. (1980). Perceptual learning a few degrees from the fovea. Farther away, specific for orientation and spatial frequency. Nature, differences are found between the vertical and London, 287, 43-44. the horizontal meridian. This suggests that, in Fiorentini, A. & Berardi, N. (1981). Learning in addition to the local orientational anisotropies, grating waveform discrimination: Specificity for orientation and spatial frequency. Vision Research, 21, which are radially symmetrical, there are func1149-1158. tional asymmetries in the representation of Fiorentini, A. & Berardi, N. (1984). Right-hemisphere supethe vertical and horizontal meridian in the riority in the discrimination of spatial phase. Perception, human visual cortex. Whether these are related 13, 695-708. to structural asymmetries such as those reported Hubel, D. H. & Freeman, D. G. (1977). Projection into the visual field of ocular dominance columns in macaque for the interhemispheric callosal connections monkey. Brain Research, 122, 336-343. (Van Essen, Newsome & Bixby, 1982) or LeVay, S., Hubel, D. H. & Wiesel, T. N. (1975). The pattern for the pattern of ocular dominance columns of ocular dominance columns in macaque visual cortex outside the parafoveal region of the revealed by a reduced silver stain. Journal of Comparative monkey striate cortex (LeVay et al., 1975; Neurology, 159, $59-576. I-MA & Freeman, 1977; Tootell, Switkes, Leventhal, A. G. & &hall, J. D. (1983). Structural basis of orientation sensitivity of cat retinal ganglion ceils. Journal Silverman & Hamilton, 1988) remains to be of Comparative Neurology, 220, 465-475. ascertained. Levick, W. R. & Thibos, L. N. (1982). Analysis of orientation bias in cat retina. Journal of Physiology, London,

Acknowledgements-We thank Professor L. Maffei and Drs D. C. Burr and M. C. Morrone for valuable comments on an earlier version of the manuscript and Professor P. Buisseret for stimulating discussion. We also thank P. Taccini for technical assistance. N. Berardi is Associate Professor at the Department of General and Environmental Physiology, University of Naples, Italy.

REFERENCES

329, 243-26 1.

McColgin, F. H. (1960). Movement thresholds in peripheral vision. Journal of the Optical Society of America, 50, 774-779.

Morrone, M. C., Burr, D. C. & Spinelli, D. (1989). Discrimination of spatial phase in central and peripheral vision. Vision Research,

29, 433-445.

Pointer, J. S. & Hess, R. F. (1989). The contrast sensitivity gradient across the human visual field: With emphasis on the low spatial frequency range. Vision Research, 29, 1133-1151.

Bauer, R. & Dow, B. M. (1989). Complementary global maps for orientation coding in upper and lower layers of the monkey’s striate cortex. Experimental Brain Research, 76, 503-509. P. J. & Banks, M. S. (1987). Sensitivity loss in odd-symmetric mechanisms underlies phase anomalies in peripheral vision. Nature, London, 326,

knnett,

873-876.

Rentschler, I. & Treutwein, B. (1985). Loss of spatial phase relationships in extrafoveal vision. Nature, London, 313, 308-3 10.

Rovamo, J. & Virsu, V. (1979). An estimation and application of the cortical magnification factor. Experimental Brain Research,

37, 495-510.

Rovamo, J., Virsu, V., Laurinen, P. & Hyvarinen, L. (1982). Resolution of gratings oriented along and across meri-

1836

Research Note

dians in peripheral vision. Investigative Ophthalmology and Yisnal Sdence, 23, 666-670.

Schah, J. F., Perry, V. H. B Leventhal, A. G. (1986). Retinal gan@ion cell dendritic fields in Old-World monkeys are oriented radially. Brain Research, 368, 18-23. Tolhurst, D. J. & Ling, L. (1988). Magnification factors and the or~anixation of the human striate cortex. Human Neurobiology, 4, 247-254. Tootell, R. B. H.. Switkes, E., Silverman, M. S. Br Hamilton, S. L. (1988). Functional anatomy of macaque striate

cortex-4

Retinotopic organixation. Journal of Neuro -

science, 8, 1531-1568.

Van Essen, D. C., Newsome, W. T. & Bixby, J. L. (1982). The pattern of interhemispheric connections and its relationship to extrastriate visual areas in the macaque monkey. Journal of Neuroscience, 2, 265-283. Yap, Y. I., Levi, D. M. & Klein, S. A. (1987). Peripheral hyperacuity: Isoeccentric bisection is better than radittl bisection. Joumal of the Optical Society of America, 4A, 1562-l 567.

Visual field asymmetries in pattern discrimination: a sign of asymmetry in cortical visual field representation?

A visual field asymmetry is described relative to the discrimination of mirror symmetric bars with ramp-like luminance profiles. Along the vertical me...
590KB Sizes 0 Downloads 0 Views