SHORT COMMUNICATIONInterocular temporal delay sensitivity in the visual cortexof the awake monkey

Rogelio Perez,1,2 Francisco Gonzalez,1,2 Maria S. Justo1 and Carlos Ulibarrena1

1Departamento de Fisiologia, Laboratorios ¢Ramon Dominguez', Facultad de Medicina, Universidad de Santiago deCompostela, Spain2Servicio de Oftalmologia, Complejo Hospitalario Universitario de Santiago, Santiago de Compostela, Spain

Keywords: binocular vision, Macaca mulatta, stereopsis, striate cortex, temporal disparity

Abstract

Due to the separation of the eyes, temporal retinal disparities are created during binocular stimulation and they have been proposedto be the basis of several stereo-visual effects. This paper studies the sensitivity of cortical neurons from area V1 to interoculartemporal delay in the awake monkey (Macaca mulatta). Forty-four cells were included in this study. Temporal delay sensitivity wasobserved in 59% of them. About half of these temporal-delay-sensitive cells were also sensitive to the stimulation sequence of theeyes. The cells that preferred one eye to be stimulated ®rst were termed asymmetrical (46%); those which were not sensitive to theeye sequence of stimulation were termed symmetrical (54%). No clear differences were observed in the distribution of delay-sensitivecells according to their eye dominance. Fifty-six percent of balanced cells and 65% of unbalanced cells were sensitive to interoculardelay. These data underline the importance of temporal cues for depth perception.

Introduction

Because of horizontal eye separation, the left and right retinal images

are slightly different. These differences are termed spatial disparities

and the visual system makes use of them to achieve depth perception,

as demonstrated by the use of random dot stereograms (Julesz, 1960).

Sensitivity to spatial disparities has been described in cortical neurons

of monkeys, cats and other animals (see Gonzalez & Perez, 1998 for

review).

Additionally, the separation of the eyes also causes temporal retinal

disparities to occur during object or observer motion. Interocular

temporal input delays are believed to be the basis of stereo-effects

such as Pulfrich's or March-Dvorak's phenomena (Howard &

Rogers, 1995). Sensitivity of cortical neurons to interocular temporal

disparities is known to exist in cats. It has been shown that cortical

visual cells in this animal are driven with unequal latencies which

could provide the substrate for a temporal disparity depth perception

mechanism (Gardner et al., 1985; Gardner & Raiten, 1986). These

observations suggest that visual neurons can use time-based cues to

calculate stereoscopic depth. Thus, sensitivity to spatial disparity and

sensitivity to interocular temporal disparity should be regarded as two

complementary mechanisms to achieve reliable depth perception. The

purpose of this study was to investigate the cell responses of visual

area V1 to binocular stimulation with different interocular stimulus

delays in awake monkeys.

Methods

The experimental procedures are described in detail elsewhere

(Gonzalez et al., 1993a,b) and are brie¯y outlined here.

One male Macaca mulatta weighing 5 kg was trained to maintain a

steady binocular ®xation on a small bright bar (0.2 3 0.1 °). The

animal was sitting with his head ®xed in front of two cold mirrors to

allow simultaneous and separate viewing of two black±white

monitors located 57.7 cm away. The stimulus was delivered during

the period of ®xation and consisted of a static square bright ®gure

(1.5 3 1.5°) on a dark background brie¯y ¯ashed (one monitor frame-

time, 20 ms) over the cell's receptive ®eld. Cells were stimulated by

presenting the stimulus over a range of interocular delays at 20-ms

steps. The temporal delays and eye sequence of stimulation were

randomized.

Cell responses were quanti®ed by subtracting the basal discharge

rate from the discharge rate (spikes per second) calculated for a

period of 150 ms after the stimulus was presented to the delayed eye.

The average of at least 15 stimulus presentations under the same

conditions was used to construct plots of cell response against

stimulus delay for each cell (Figs 1 and 2). In order to assess

sensitivity to interocular temporal delay, and the symmetry of the

responses relative to zero delay, the analysis of variance (ANOVA,

P < 0.05) was used.

Monocular stimulus presentations lasting 750 ms were used to

assess the cell's ocular dominance. Eye dominance was de®ned by the

eye dominance index which resulted from dividing the cell response

evoked from the contralateral eye by the sum of the responses obtained

from the contra- and ipsilateral eyes. According to the eye dominance

index, the cells were organized in seven groups (Hubel & Wiesel,

1962, 1968). Cells belonging to groups 1, 2, 6 and 7 were considered

Correspondence: Dr Francisco Gonzalez, Departamento de Fisiologia,Facultad de Medicina, E-15705 Santiago de Compostela, Spain.E-mail: fspaco@usc.es

Received 25 January 1999, revised 19 April 1999, accepted 20 April 1999

European Journal of Neuroscience, Vol. 11, pp. 2593±2595, 1999 ã European Neuroscience Association

unbalanced cells, whereas those included in groups 3, 4 and 5 were

considered balanced cells.

To access the cortex with metal electrodes, several craniotomies

were performed on the skull covering the occipital lobe under

ketamine anaesthesia (7 mg/kg, i.m.). For major surgical procedures

the animal was anaesthetized with sodium pentobarbital (27 mg/kg,

i.v.) after induction with ketamine (25 mg, i.m.). Antibiotics

(Penicillin, 50 000 IU/kg, i.m.) and analgesics (Noramidopirine,

150 mg/kg, i.m.) were given after surgery. All efforts were made to

minimize animal suffering and the surgical and experimental

procedures followed the guidelines of the Bioethic Committee of

our institution.

Results

This study was conducted in 44 cells recorded from the dorsal and

dorsomedial surface of area V1. The eccentricity of the receptive ®eld

positions ranged from 6.6 to 18.1 ° (mean, 10.9 °). Sixty-one percent

of cells (27 out of 44) were ocular-balanced while 39% (17 out of 44)

were considered ocular-unbalanced cells.

Slightly more than half of recorded neurons (26 out of 44, 59%)

showed different responses to different interocular delays and

therefore they were termed delay-sensitive cells. Typically, delay-

sensitive units (Figs 1 and 2) displayed an interocular facilitation for a

given range of delays, and a relatively ¯at response pro®le for delays

exceeding this range. This facilitation was centred at, or around, zero

interocular delay, within the range of 6 100 ms and lasted from 100

to 200 ms in most cells (ANOVA, P < 0.05).

In about half of delay-sensitive cells (12 out of 26, 46%) there was

an increased response away from zero interocular delay, indicating

that the cell was sensitive to the sequence of eye stimulation (Fig. 1).

In these cells, termed asymmetrical delay-sensitive cells, the

responses obtained under left±right and right±left sequences of eye

stimulation were statistically different (ANOVA, P < 0.05). In the

remaining delay-sensitive cells (14 out of 26, 54%) there was an

increased response centred at zero interocular delay, showing no

signi®cant statistical differences regarding the eye sequence of

stimulation (ANOVA, P > 0.05). These cells were termed symmetrical

delay-sensitive cells (Fig. 2). We were unable to ®nd any relationship

between eye dominance and these two types of cells (asymmetrical

and symmetrical).

Delay-sensitive cells were in quite similar percentages from both

balanced (15 out of 26, 58%) and unbalanced (11 out of 26, 42%)

groups. In contrast, delay-insensitive cells were twice as frequently

balanced (12 out of 18, 67%) than unbalanced (six out of 18, 33%).

Discussion

Psychophysical observations suggest that the visual system is

sensitive to temporal disparities. It was shown that a target is

perceived beyond the ®xation point when it is ¯ashed ®rst to one eye

and then to the other with a temporal offset, and that the distance

increases with increasing temporal offset (Wist, 1968). Moreover,

interocular delay elicited by stroboscopic stimulation has been shown

to be a potent source for depth perception (Burr & Ross, 1979).

There is also experimental evidence that cortical visual cells are

capable of signalling interocular temporal disparities. It has been

observed that some cortical cells of the cat modulate their response to

spatial disparity when interocular delay is added to the stimulus

(Pettigrew et al., 1968) and that they are selectively tuned to both

spatial and temporal disparities (Cynader et al., 1978; Gardner et al.,

1985). Interocular temporal delay generated by placing a ®lter before

one eye was associated with a shift in the tuning of cortical cells to

spatial disparity in cats, similar in magnitude to the Pulfrich effect in

humans (Carney et al., 1989). In agreement with these ®ndings, the

present study shows that there is a population of cells in monkey

visual area V1 sensitive to interocular stimulus delay.

The delay-sensitive cells we found are able to encode interocular

temporal disparity; this could be used to detect object trajectories in

depth. Indeed, a target which moves in front or behind the ®xation

point from left to right will stimulate a given corresponding retinal

area ®rst in one eye and then in the other eye. The delay will be a

function of the speed and trajectory of the target relative to the

®xation point. For cells with their left and right receptive ®elds in

register, a null delay would indicate that the target is on the horopter,

FIG. 1. Visual responses of a cell recorded from area V1 of an awake monkeyto a binocularly, brie¯y ¯ashed stimulus (20 ms) with different interoculardelays. Each dot represents the difference between the stimulus response to thesecond eye stimulus and the basal activity of the cell (2.5 spikes/s). Thevertical bars represent the SD (see methods). Negative delays indicate that theright eye was stimulated ®rst. Positive values indicate that the left eye wasstimulated ®rst. The cell optimal response is slightly shifted towards negativedelays (ANOVA, P < 0.05). This cell was considered an asymmetrical delay-sensitive cell. Monocular responses to the left (L, ®lled circle) and right eye(R, triangle) stimuli are also represented.

FIG. 2. Responses of a visual cell of monkey area V1 to a binocularly, brie¯y¯ashed stimulus (20 ms) with different interocular delays. The plot wasconstructed as in Fig. 1. The basal activity of the cell was 13.5 spikes/s. Thecell displays and enhancement of the response to the stimulation of the secondeye for delays ranging from ±100 to +100 ms (ANOVA, P < 0.05). This cell wasconsidered a symmetrical delay-sensitive cell.

2594 R Perez et al.

Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 2593±2595

whereas a given delay would indicate that the target is moving in

front or behind the horopter. There is evidence that cortical cells may

have left and right receptive ®elds which are not in register in

monkeys (Hubel & Wiesel, 1970) and cats (Barlow et al., 1967; von

der Heydt et al., 1978; Ferster, 1981; Maske et al., 1984; Bishop &

Pettigrew, 1986). For these cells with disparate receptive ®elds a null

delay would indicate the target is moving on a plane in front or

behind the horopter matching their receptive ®eld disparity, whereas a

given delay would indicate the target is in front or behind this plane.

The cells we have termed symmetrical are unable to provide

information about the direction of motion of the object. In contrast,

asymmetrical cells would signal object trajectories off the plane

matching their receptive ®elds and would be able to detect which eye

was stimulated ®rst. Therefore they provide information about the

direction of motion. It is clear however, that additional information

on the position of the object relative to the horopter is needed to

de®ne the direction of movement fully.

The responses of symmetrical and asymmetrical cells are

consistent with the ®nding that some cortical visual cells in cats

may have equal or unequal latencies with the two eyes (Gardner et al.,

1985). If a neuron has facilitatory interocular interaction, is activated

with equal latencies from both eyes and its receptive ®elds are located

on corresponding retinal points, then the cell would be maximally

activated when the stimulus is on the horopter. This would be the case

with symmetrical cells. On the other hand, if one of the two eyes has

longer latency, an optimal response would be evoked when the

stimulus is moving in front or beyond the horopter, crossing the left

and right visual axis with a delay matching the eye latency. This

would be the case with asymmetrical cells.

The response pro®le of symmetrical cells was centred at zero

delay, whereas that of asymmetrical cells was centred at positive or

negative delays. The half-width of the temporal selectivity curves

varied from cell to cell, having an average duration of » 150 ms.

Although the delay between the stimulation of two corresponding

retinal points depends on the speed and trajectory of the stimulus, the

interaction time characteristics we observed suggest that the

sensitivity to interocular delay is more appropriate for detection of

trajectories of objects moving slowly on or around the horopter

(probably restricted to the Panum's area) or fast moving objects

moving off the horopter.

Our data indicate that the visual system is able to monitor the

temporal difference in the stimulation of points on the two retinas

which could account for depth perception and stereopsis. This

®nding emphasizes the similarities among different sensory

systems. For instance, experiments made in hearing showed that

small time differences between the stimulation of the left and

right ear induced a spatial shift in sound perception. Similar

results were obtained for taste stimulation on both sides of the

tongue and with odourous substances introduced to both nostrils

(von Bekesy, 1969). It must be pointed out however, that

interocular temporal disparity, by itself, does not provide enough

information to compute the trajectory of an object moving in

depth. However, the information available from both types of

delay-sensitive cells we have described here can be completed

with additional information about speed, direction of movement

and spatial disparity of the stimulus. There is currently plenty of

evidence that these parameters are accurately encoded by cortical

visual cells (Orban, 1991; Gonzalez & Perez, 1998).

Acknowledgements

This work was supported by grant PB97±0521 (DGESIC-PGC) from theSpanish Ministerio de Educacion y Cultura.

References

Barlow, H.B., Blakemore, C. & Pettigrew, J.D. (1967) The neural mechanismof binocular depth discrimination. J. Physiol. (Lond.), 193, 327±342.

von Bekesy, G. (1969) The smallest difference the eyes can detect withsweeping stimulation. Proc. Natl Acad. Sci. USA, 64, 142±147.

Bishop, P.O. & Pettigrew, J.D. (1986) Neural mechanisms of binocular vision.Vision Res., 26, 1587±1600.

Burr, D.C. & Ross, J. (1979) How does binocular delay give information aboutdepth? Vision Res., 19, 523±532.

Carney, T., Paradiso, M.A. & Freeman, R.D. (1989) A physiological correlateof the Pulfrich effect in cortical neurons of the cat. Vision Res., 29, 155±165.

Cynader, M., Gardner, J. & Douglas, R. (1978) Neural mechanisms underlyingstereoscopic depth perception in cat visual cortex. In Cool, S.J. & Smith,E.L. (eds.), Frontiers in Visual Science. Springer, Berlin, pp. 373±386.

Ferster, D. (1981) A comparison of binocular depth mechanisms in areas 17and 18 of the cat visual cortex. J. Physiol., 311, 623±655.

Gardner, J.C., Douglas, R.M. & Cynader, M.S. (1985) A time-basedstereoscopic depth mechanism in the visual cortex. Brain Res., 328, 154±157.

Gardner, J.C. & Raiten, E.J. (1986) Ocular dominance and disparity-sensitivity: why there are cells in the visual cortex driven unequally bythe two eyes. Exp. Brain Res., 64, 505±514.

Gonzalez, F., Krause, F., Perez, R., Alonso, J.M. & AcunÄa, C. (1993a) Cellresponses to vertical disparity in the monkey visual cortex. Neurosci. Lett.,160, 167±170.

Gonzalez, F. & Perez, R. (1998) Neural mechanisms underlying stereoscopicvision. Prog. Neurobiol., 55, 191±224.

Gonzalez, F., Relova, J.L., Perez, R., AcunÄa, C. & Alonso, J.M. (1993b)Binocular matching in monkey visual cortex: single cell responses tocorrelated and uncorrelated stereograms. Neuroscience, 52, 933±939.

von der Heydt, R., Adorjani, C., Hanny, P. & Baumgardtner, G. (1978)Disparity-sensitivity and receptive ®eld incongruity of units in the cat striatecortex. Exp. Brain Res., 31, 523±545.

Howard, I.P. & Rogers, B.J. (1995) Binocular vision and stereopsis. OxfordPsychology Series No. 29. Oxford University Press, Oxford, UK.

Hubel, D.H. & Wiesel, T.N. (1962) Receptive ®elds, binocular interactionsand functional architecture in the cat's visual cortex. J. Physiol. (Lond.),160, 106±154.

Hubel, D.H. & Wiesel, T.N. (1968) Receptive ®elds and functionalarchitecture of monkey striate cortex. J. Physiol. (Lond.), 195, 215±243.

Hubel, D.H. & Wiesel, T.N. (1970) Cells sensitive to binocular depth in area18 of the macaque monkey cortex. Nature, 225, 41±42.

Julesz, B. (1960) Binocular depth perception of computer generated patterns.Bell Systems Techn. J., 39, 1125±1162.

Maske, R., Yamanee, S. & Bishop, P.O. (1984) Binocular simple cells forlocal stereopsis: comparison of receptive ®eld organizations of the two eyes.Vision Res., 24, 1921±1929.

Orban, G. (1991) Quantitative electrophysiology of visual cortical neurones.In Leventhal, A.G. (ed.), Vision and Visual Dysfunction. The Neural Basisof Visual Function, Vol. 4. Macmillan Press, London, UK, pp. 173±222.

Pettigrew, J.D., Nikara, T. & Bishop, P.O. (1968) Binocular interaction onsingle units in cat striate cortex: simultaneous stimulation by single movingslit with receptive ®elds in correspondence. Exp. Brain Res., 6, 391±410.

Wist, E.R. (1968) The in¯uence of the equidistance tendency on depth shiftsresulting from an interocular delay in stimulation. PerceptionPsychophysics, 8, 89±92.

Sensitivity to interocular temporal delay in macaque V1 2595

Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 2593±2595