Binocular Summation of Detection and Resolution Thresholds in the Central Visual Field Using Parallel-Line Targets

Abstract

purpose. To investigate the effect of thresholds on binocular summation for detection and resolution of parallel-line targets of different widths.

methods. The automatic perimeter Octopus 201 (Haag-Streit International, Köniz, Switzerland) combined with a modified haploscope device and targets consisting of parallel lines that were 10′ (10′ target), 2.5′ (2.5′ target), or 1.43′ (1.43′ target) of visual angle in width and apart were used to measure the thresholds for monocular–binocular detection and resolution of the targets under the same binocular fusion stimulation conditions in seven young adults with normal vision. The custom program in the automatic perimeter was used to test 27 points. Seventeen of the 27 points were located in the central 6° of visual field and 10 on the horizontal meridian subtending visual angles of 8°, 10°, 12°, 16°, and 20°.

results. The resolution threshold was significantly higher than the detection threshold for the 2.5′ and 1.43′ targets (P < 0.01 by Bonferroni/Dunn test). Furthermore, as the target width decreased to 2.5′ or 1.43′, the binocular summation ratio for the resolution threshold increased significantly over that for the detection threshold, with increasing distance from the fovea (P < 0.01 by Wilcoxon signed ranks test).

conclusions. Binocular summation for detection and resolution thresholds varies as a function of the width of a parallel-line target. The difference between binocular summation ratios for detection and resolution thresholds increases with decrease in target width and increase in eccentricity from the fovea. Binocular interaction plays an important role in the task of recognizing a high-resolution target in the parafoveal area.

Binocular summation is defined as an increase in binocular performance over monocular performance. The visual information received from both eyes converges on and is processed by the binocular cells in the visual cortex. The proportion of binocular cells in the visual cortex that are excited when stimulation is received from both eyes has been reported to be between 70% and 80%. ¹ These binocular cells are excited when the right and left receptive fields exactly superimpose. ^{2 3} Therefore, when measuring visual sensitivity to evaluate binocular summation, it is essential to confirm the condition of binocular vision—that is, the point at which vision fuses during binocular processing. For this purpose, we need a device to confirm whether the binocular performance is binocular fusion or suppression. Furthermore, when comparing monocular and binocular visual sensitivities, visual sensitivities should be measured within the same binocular performance, and measurement of monocular visual sensitivity should be taken without occluding the nontested eye. However, to our knowledge, no study has yet been reported in which binocular summation is examined by measurements of vision in the right and left eyes separately and together, without occluding the nontested eye during presentation of a binocular fusion stimulus. We therefore developed an apparatus by installing a binocular fusion stimulus device into an automatic perimeter. It enabled us to confirm the conditions of binocular vision when measuring the visual sensitivities of the right and left eyes separately and together. ^{4 5}  

Various factors affect binocular summation. For example, when flicker and grating stimuli are used, the amount of binocular summation for flicker and grating contrast sensitivities decreases as the nasotemporal asymmetry increases. ⁶ When orientation-discrimination stimuli are used, binocular summation is found to be dependent on stimulus contrast and duration. ⁷ In another study reporting the influence of contrasts on binocular summation in vernier acuity, significant summation has been found for low-contrast stimuli but negligible summation for high-contrast stimuli. ⁸ As for the influence of stimulus size, Wood et al. ⁹ evaluated binocular summation in the fovea and peripheral visual field and found that binocular summation in the periphery decreases with decreasing stimulus size and increases with increasing stimulus size. In our previous study on the influence of stimulus size, we evaluated binocular summation in the central 6° of the visual field with targets of six different sizes and showed the highest degree of binocular summation for the smallest target size (0.054°) used but no further decrease in binocular summation for target sizes larger than 0.108° (up to 1.724°). ⁴ We have also found that binocular summation varies according to the area of the retina that is stimulated. ⁴  

Zlatkova et al. ¹⁰ examined the differences between binocular summation for grating detection and resolution at the fovea and in the periphery using grating targets of various spatial frequencies. They found no significant difference in binocular summation for detection and resolution acuity at the fovea, but they found an increase in resolution acuity at 25° in the inferior field. Many previous studies of binocular summation have focused on resolution and detection thresholds in the fovea and the periphery. However, whether there is a difference between detection and resolution thresholds within the parafoveal area is still unclear. If it does, its association with various locations in the parafoveal area should be examined. Clarification of these questions can provide more information on how binocular summation performs for resolution task in the parafoveal area. Thus, we conducted the present study to elucidate the differences in binocular summation for detection and resolution thresholds in the central 6° of the visual field. 

Materials and Methods

Subjects

Subjects were seven individuals between 22 and 26 years of age without any systemic or ophthalmic diseases likely to alter their visual function. For this study, we selected the subjects who were proficient in psychophysical experiments. The inclusion criteria were as follows: corrected visual acuity of 1.0 or better, refractive error within 1.0 D of spherical error and within 0.75 D of astigmatism, and binocular function of 60 seconds of arc or better on the TNO stereo test. 

All experiments were performed in accordance with the Declaration of Helsinki for research involving human subjects. Informed consent was obtained from all subjects after explanation of the nature and possible consequences of the study. 

Apparatus and Stimulus

When measuring detection and resolution thresholds, we used the Octopus 201 perimeter (Haag-Streit International, Köniz, Switzerland) combined with the space synoptophore (a modified haploscope device) developed by Iwai ¹¹ (Fig. 1) . In this modified device, the refraction mirrors of the standard haploscope were replaced with half mirrors. We used the space synoptophore to confirm the binocular condition by projecting fusion patterns on the front of the half mirrors. The fusion patterns used were a 3° square for each eye, a 135° line for the left eye, and a 45° line for the right eye (Fig. 2) . To project the fusion patterns, we removed the projection lamp of the space synoptophore and used the background luminance of the Octopus 201 perimeter. A parallel-line target was then projected on the cupola of the Octopus 201 perimeter with a background luminance of 4 apostilb (asb) and stimulus duration of 100 ms. Thresholds were determined using a 4-2-1 dB bracketing staircase measurement procedure. Three parallel-line targets with various widths and distances between the lines of 10′, 2.5′, or 1.43′ angle of vision and a length of 1.724° were used (Fig. 2) . This system enabled us to confirm the conditions of binocular vision when measuring the visual sensitivities of the right and left eyes separately and together (Fig. 1) . Fixation was checked in two ways. One was for the examiner to monitor the pupils’ positions of both eyes with a small infrared video camera. The measurement was to be continued only when the subject fixed the central fixation point within the cupola of the Octopus 201 under the binocular condition. The other was for the subject to push the button continuously. The subjects were instructed in advance to discontinue the measurement by continuously pushing the button as soon as the fusion patterns were perceived to be broken. In other words, the measurement would be discontinued if the subject’s fixation had shifted or if the subject had continuously pushed the button. 

The Sargon program is designed to project any test point in the Octopus 201 perimeter. With this program, we tested 27 points. Among the 27 points, 17 were located in the central 6° visual field and 10 on the horizontal meridian subtending 8°, 10°, 12°, 16°, and 20°. 

Measurement Procedures

We measured detection and resolution thresholds in the right and left eyes separately and together under the same binocular fusion stimulation conditions. The same fusion patterns were presented to all subjects. The detection threshold was evaluated when the subject perceived the target. The resolution threshold was evaluated when the subject recognized the parallel-line pattern of the target. The resolution threshold was defined as the lowest contrast at which two parallel lines were perceived. To measure monocular visual sensitivity under binocular vision, the nontested eye was occluded with a cover of a color similar to that of the cupola (Fig. 3) . Thus, the fusion patterns were presented to both eyes, but the target stimulus was blocked from the nontested eye. The subject did not notice which eye was being tested. The subject’s perception of the target was identified when the subject pressed the hold button. 

Three measurements of visual sensitivity were obtained for each target and test condition in each subject. The first measurement in each set of three was excluded, and the average of the second and third measurements was used for all subsequent calculations for that set. The order of eyes to be tested and the order for presenting targets of various widths were determined randomly for each subject. 

Results

Visual Sensitivities for Detection and Resolution

The visual sensitivities for both detection and resolution were higher in both eyes together than in the individual eyes for all target widths (P < 0.01; Bonferroni/Dunn tests; Table 1 ). The visual sensitivity for resolution for both eyes together exceeded that predicted by probability summation. 

All targets could be detected within the central 20° of the visual field. However, for resolution task, the 2.5′ and 1.43′ targets could not be perceived in the periphery. As the target became narrower, the area of the visual field over which it could be perceived became more centralized. Specifically, the 2.5′ target was resolved only within the central 8° of the visual field and the 1.43′ target could only be perceived within the central 4° of the visual field (Fig. 4) . 

Differences between Detection and Resolution Thresholds

Resolution thresholds were higher than detection thresholds for all target widths (P < 0.01; Wilcoxon signed ranks test). Figure 5shows the difference in threshold energy between detection and resolution thresholds within the central 6° of the visual field. Threshold energy is calculated as log (L + ΔL) × A, where L is the background luminance; ΔL, the stimulus luminance; and A, the stimulus area. The differences between values for the 2.5′ and 1.43′ targets were significant for either the right or left eye versus both eyes (P < 0.01; Bonferroni/Dunn test; Fig. 5 ). 

Binocular Summation Ratio

Figure 6shows the binocular summation ratios (binocular/monocular threshold) for detection and resolution of targets 10′, 2.5′, or 1.43′ wide at various eccentricities within the center 6° of the visual field. 

The binocular summation ratios for detection and resolution thresholds were not significantly different for the 10′ target at any distance from the center of the visual field. However, for the 2.5′ and 1.43′ targets, the binocular summation ratios for resolution thresholds were significantly higher than the ratios for detection thresholds (P < 0.01; Wilcoxon signed ranks test). Besides, the differences increased as the target width decreased—that is, the ratio was highest for the 1.43′ targets for each stimulus area in the central 6° of the visual field. Moreover, the ratio increased as the stimulus was presented farther away (i.e., with increasing eccentricity from the fovea). 

Discussion

This study clearly showed that the binocular summation ratio for resolution threshold was significantly higher than that for detection threshold as the target width and the distance between parallel lines decreased. This result indicates that binocular processing is necessary when resolving a high-resolution target. In addition, increasing eccentricity within the central 6° of the visual field was accompanied by an increase in the binocular summation for resolution threshold and also by an increase in the difference between binocular summations for resolution and detection thresholds. This finding suggests that for the consecutive processing of visual information that is received binocularly, resolution of the image by binocular summation in the parafoveal area facilitates resolution of the image at the fovea. As shown in our previous results, the size of the receptive field for binocular stimulation under conditions requiring binocular fusion is smaller than that for monocular stimulation under the same conditions. ⁴ The difference between detection and resolution thresholds was found to be smaller in binocular conditions than in monocular conditions. Therefore, a possible explanation is that resolution becomes higher as the size of the receptive field under binocular condition decreases. 

There have been many studies focused on either the fovea or the periphery. Some studies of the fovea indicated that neither contrast levels nor orientation of the target affects binocular summation. ¹² Zlatkova et al. ¹⁰ showed that there are no significant differences between detection and resolution acuity at the fovea. In our present study, the difference between binocular summation for detection and resolution thresholds was found to be smaller at the fovea than in other areas. This finding corresponds to their result. As for vision in the periphery, Zlatkova et al. also reported a significant difference between binocular summation for detection and resolution acuity at 25° in the inferior field and particularly a large improvement in resolution acuity when viewing binocularly. Pardhan ¹² found that at 8° eccentric to the fovea, binocular summation is higher when measured with a low-contrast grating than with a high-contrast grating. By Grisby and Tsou, ⁶ with an emphasis on the relationship between binocular summation and nasotemporal asymmetries in the periphery, significant binocular summation was found at 4° and 8°, and its level declined with the increasing nasotemporal asymmetries beyond 20° eccentricity in the periphery. However, no nasotemporal asymmetry was found in visual sensitivities within the central 6° of the visual field, ¹³ and this is an important consideration for our testing conditions in this study. 

When assessing the difference between monocular and binocular performance in the periphery, eye–field asymmetry is an problem to be concerned about. Zackon et al. ¹⁴ reported a subcortical attentional effect. They pointed out that the perceived collision point is closer to fixation when the initial cue is presented to the left eye for monocular presentation whereas the right eye yields the same results as those of binocular presentation. Fujimoto and Adachi-Usami ¹⁵ reported that visual sensitivity increases with a decrease in the size of test field and with the diminished number of test points. Prediction and attention are thought to cause the increase in sensitivity. In our tests, to minimize the visual sensitivity variation due to visual attention or prediction, we tested 27 points that were widely distributed on the horizontal meridian up to 20°, but we focused our studies on the parafoveal area where no significant differences were found in visual sensitivity between the nasal retina (temporal field) and the temporal retina (nasal field) in the right or left eye. Furthermore, when a stimulus target is randomly projected, the subject cannot predict its position and is therefore less influenced by the visual attention or prediction. Thus, we consider our test conditions in this study adequate for the assessment of the difference in binocular summation for resolution and detection thresholds. 

In conclusion, we demonstrated that the thresholds for resolution and detection of binocular fusion stimuli can be significantly different, depending on the width and the distance between parallel lines and on the location in the parafoveal area where the thresholds are measured. Along with increasing eccentricity from the fovea, the difference in binocular summation ratio between detection and resolution thresholds also increases. In future studies, it is our intent to investigate further the binocular summation in the foveal and parafoveal areas in patients with ocular diseases. 

 Supported by Grant 16922203 from the Japanese Ministry of Education.

 Submitted for publication December 5, 2004; revised March 30, 2005; accepted April 20, 2005.

 Disclosure: A. Wakayama, None; C. Matsumoto, None; Y. Shimomura, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Schematic representation of the system used to measure binocular summation for detection and resolution thresholds. Key components shown include the cupola of the Octopus 201 perimeter (A; Haag-Streit International, Köniz, Switzerland), the space synoptophore (a modified haploscope device), in which the refraction mirrors of the standard haploscope device were replaced with half mirrors (B), and the half-mirrors on which the fusion patterns were projected (C). Dark gray hatched area: binocular visual field, which is the overlapped section of the right and left visual fields (light gray hatched areas). Visual sensitivities were measured only within the binocular visual field. Confirmation of the subject’s viewing the fusion patterns would ensure binocular performance while measuring visual sensitivity.

Figure 2.

 

The fusion patterns used were a 3° square for each eye, a 135° line for the left eye, and a 45° line for the right eye. They were projected by the space synoptophore. Three parallel-line targets with various widths and distances between the lines of 10′, 2.5′, or 1.43′ angle of vision and a length of 1.724° were used. The parallel-line targets were projected by the perimeter.

Figure 3.

 

Schematic representation of the measurements for binocular and monocular visual sensitivities under the same binocular fusion conditions. When measuring monocular visual sensitivity under binocular vision, the nontested eye was occluded with the cover of a color similar to that of the cupola. The fusion patterns were presented to both eyes, but the parallel-line target was only to the tested eye. The subject did not notice which eye was being tested.

Table 1.

 

Binocular Summation for Detection and Resolution Visual Sensitivities for Three Widths of Parallel-Line Target

Threshold	10′			2.5′			1.43′
		Eye, Pts	dB	Eye, Pts	dB	Eye, Pts	dB
Detection	R, 25	32.8 ± 1.1	R, 25	29.9 ± 1.2	R, 25	28.9 ± 1.1
	L, 25	32.7 ± 1.0	L, 25	30.1 ± 1.0	L, 25	29.0 ± 1.2
	B, 27	34.1 ± 1.2	B, 27	31.2 ± 1.3	B, 27	30.1 ± 1.4
	P < 0.001			P < 0.001			P < 0.001
			R, 19	30.3 ± 0.9	R, 11	29.6 ± 0.8
			L, 19	30.4 ± 0.8	L, 11	29.8 ± 0.6
			B, 19	31.7 ± 1.0	B, 11	31.1 ± 1.0
			P < 0.001			P < 0.001
Resolution	R, 25	32.0 ± 0.9	R, 19	28.4 ± 1.2	R, 11	26.8 ± 1.5
	L, 25	32.0 ± 1.0	L, 19	28.1 ± 1.2	R, 11	26.7 ± 1.1
	B, 27	33.5 ± 1.1	B, 19	30.0 ± 1.1	B, 11	29.1 ± 0.9
	P < 0.001			P < 0.001			P < 0.001

 Data are the mean ± SD for the second and third measurements in each of the seven subjects. Probabilities are for the differences in thresholds for monocular versus binocular vision (Kruskal-Wallis and Bonferroni/Dunn tests). Pts, test points possible to be measured in the visual field, of a total of 27 (17 in the central 6° of the visual field and 10 on the horizontal meridian subtending angles of 8°, 10°, 12°, 16°, and 20°); R, right-eye binocular vision; L, left-eye binocular vision; B, both eyes in binocular vision.

Figure 4.

 

Visual sensitivities (mean ± SD) for detection and resolution under conditions of the right or left eye binocular vision or both eyes together along the horizontal meridian (eccentricity) encompassing the central 6° of the visual field and up to 20° on either side. The y-axis represents visual sensitivity and the x-axis, eccentricity from the fovea.

Figure 5.

 

For the 2.5′ and 1.43′ target in the central 6° of visual field, resolution thresholds were significantly higher than detection thresholds for the right and left eyes separately or together (P < 0.01; Bonferroni/Dunn test). The x-axis represents eccentricity and the y-axis, the difference between resolution and detection thresholds.

Figure 6.

 

Binocular summation ratio as a function of eccentricity within the central 6° of the visual field. For the 2.5′ target versus 1.43′ target, binocular summation for resolution was significantly higher than that for detection. With increasing distance from the fovea, the differences in binocular summation between resolution and detection also significantly increased (P < 0.01; Wilcoxon signed-ranks test). The y-axis represents the binocular summation ratio (binocular/monocular thresholds) and the x-axis, the width of the parallel-line target.