Distance Stereotest Using a 3-Dimensional Monitor for Adult Subjects JONGSHIN KIM, HEE KYUNG YANG, YOUNGMIN KIM, BYOUNGHO LEE, AND JEONG-MIN HWANG ● PURPOSE:

To evaluate the validity and test–retest reliability of a contour-based 3-dimensional (3-D) monitor distance stereotest (distance 3-D stereotest) and to measure the maximum horizontal disparity that can be fused with disparity vergence for determining the largest measurable disparity of true stereopsis. ● DESIGN: Observational case series. ● METHODS: Sixty-four normal adult subjects (age range, 23 to 39 years) were recruited. Contour-based circles (crossed disparity, 5000 to 20 seconds of arc; Microsoft Visual Studio Cⴙⴙ 6.0; Microsoft, Inc, Seattle, Washington, USA) were generated on a 3-D monitor (46-inch stereoscopic display) using polarization glasses and were presented to subjects with normal binocularity at 3 m. While the position of the stimulus changed among 4 possible locations, the subjects were instructed to press the corresponding position of the stimulus on a keypad. The results with the new distance 3-D stereotest were compared with those from the distance Randot stereotest. ● RESULTS: The results of the distance 3-D stereotest and the distance Randot stereotests were identical in 64% and within 1 disparity level in 97% of normal adults. Scores obtained with the 2 tests showed a statistically significant correlation (r ⴝ 0.324, P ⴝ .009). The half-width of the 95% limit of agreement was 0.47 log seconds of arc (1.55 octaves) using the distance 3-D stereotest—similar to or better than that obtained with conventional distance stereotests. The maximum binocular disparity that can be fused with vergence was 1828 ⴞ 794 seconds of arc (range, 4000 to 500). ● CONCLUSIONS: The distance 3-D stereotest showed good concordance with the distance Randot stereotest and relatively good test–retest reliability, supporting the validity of the distance 3-D stereotest. The normative data set obtained from the present study can serve as a useful reference for quantitative assessment of a wide range of binocular sensory abnormalities. (Am J OphAccepted for publication Sep 24, 2010. From the Department of Ophthalmology, Seoul National University College of Medicine, Seoul, Korea (J.K., H.K.Y., J.-M.H.); the Department of Ophthalmology, Seoul National University Bundang Hospital, Seongnam, Korea (H.K.Y., J.-M.H.); and the School of Electrical Engineering, Seoul National University, Seoul, Korea (Y.K., B.L.). Inquiries to Jeong-Min Hwang, Department of Ophthalmology, Seoul National University Bundang Hospital, 166 Gumiro Bundang-gu Seongnam Gyeonggi-do 463-707, Korea; e-mail: [email protected] 0002-9394/$36.00 doi:10.1016/j.ajo.2010.09.034

©

2011 BY

thalmol 2011;151:1081–1086. © 2011 by Elsevier Inc. All rights reserved.)

D

ISTANCE STEREOACUITY CAN BE MEASURED WITH

the distance Randot stereotest (Stereo Optical Co, Inc., Chicago, Illinois, USA) or Frisby-Davis distance stereotest (Frisby Stereotest, Sheffield, United Kingdom). However, there are some limitations to these stereotests. First, these stereotests cannot measure stereoacuity worse than 400 seconds of arc (arcsec).1,2 Second, the predefined intervals of these distance stereotests are less compact in the range of small disparities (ⱕ 400 arcsec) than that of near stereotests such as the Randot Preschool stereotest or Frisby near stereotest. Third, these tests consist of a predefined answer that could cause learning effects if repeated several times.3 Finally, their results are affected by the state of illumination in the examination room4 because they have no self-emitting devices. The introduction of a 3-dimensional (3-D) stereoscopic monitor enables clinicians to evaluate stereopsis using more variable stereograms and more quantitative methods. Fujikado and associates used a 3-D monitorbased stereotest to evaluate stereopsis in children with strabismus.5 Breyer and associates6 developed a 3-D monitor-based stereotest to test stereovision in preverbal children. However, these studies were conducted with a small 18-inch 3-D monitor that has relatively low brightness and resolution (100 cd/m2, 640 ⫻ 480 pixels; 58 cd/m2, 640 ⫻ 1024 pixels; respectively). This limits its ability to display a broad range of disparities with desirable luminance to perform general-purpose tasks. Furthermore, this monitor has a short viewing distance that cannot measure distance stereoacuity. In this study, we developed a contour-based stereotest using a large 46-inch, high-brightness, and high-resolution 3-D monitor, which can measure a wide range (5000 to 20 arcsec) of distance stereopsis with more compact intervals in the range of small disparities. Using this distance 3-D stereotest, we measured the distance stereoacuity of normal young adults. We compared the results from the distance 3-D stereotest with those from the distance Randot stereotest and examined its test–retest reliability. Finally, we measured the maximum horizontal disparity that can be fused with disparity vergence to determine the largest measurable disparity of true stereopsis in normal subjects using this distance 3-D stereotest.

ELSEVIER INC. ALL

RIGHTS RESERVED.

1081

● CONTOUR-BASED

CIRCLE STEREOGRAM: Contourbased circles with a diameter of 14 cm and a wide range of crossed horizontal disparities (5000 to 20 arcsec) were generated on the 3-D monitor. Four circles were placed at the edges of a square with a distance of 16 cm between circles. The inside and outside of circles were filled with random dot textures to eliminate monocular cues (Figure 1). The circle stimuli were programmed in Visual C⫹⫹ using Visual Studio (version 6.0; Microsoft, Inc., Seattle, Washington, USA). Subjects viewed the stereogram at a distance of 3 m while wearing polarizing glasses. The room illumination was recorded as 250 to 300 lux, measured in front of the eye level of the subject by an illuminance meter (T-10; Konica Minolta, Tokyo, Japan). While the position of the stereoscopic stimulus was changed among the 4 possible locations, the subjects were instructed to press the corresponding position of the stimulus on a keypad. Testing commenced with coarse disparities that progressively became smaller as follows: 5000, 4000, 3500, 3000, 2500, 2000, 1500, 1000, 800, 600, 500, 400, 300, 200, 150, 100, 70, 60, 50, 40, 30, 25, and 20 arcsec. We measured the stereoacuity threshold and the maximum binocular horizontal disparity that could be fused with disparity vergence with the distance 3-D stereotest. The stereoacuity threshold was defined as the smallest disparity level at which a subject could detect a protruded circle; the maximum binocular horizontal disparity that could be fused with vergence was defined as the largest disparity level at which a subject could perceive the stereogram as a single circle. Because of the possibility of monocular cues providing false-positive responses at thresholds of 400 to 20 arcsec, those levels were reassessed monocularly. This monocular testing phase was carried out after the threshold had been reached and was performed to ensure that the threshold score reflected a binocular response. In our study, no patients who achieved a threshold of 400 to 20 arcsec were able to achieve the same levels monocularly. Each test was administered following a standardized testing protocol. The results of this distance 3-D stereotest were compared with those of the distance Randot stereotest.

FIGURE 1. Photograph showing a subject wearing polarized glasses and looking at the contour-based circle stereogram presented on a 46-inch polarized stereoscopic monitor at a distance of 3 m. Circles with 14-cm diameters and a wide range of crossed disparities were generated in 4 different locations. The insides and outsides of circles were filled with random dot textures to eliminate monocular cues. The position of the stimulus was changed among 4 possible locations, and the subjects were instructed to press the corresponding position of the stimulus on a keypad.

METHODS ● PATIENTS: Normative data for the distance Randot stereotest were obtained from 64 volunteers with normal binocularity (mean age ⫾ standard deviation, 30.7 ⫾ 4.1 years; age range, 23 to 39 years). Informed consent was obtained from all participants after the details of the study were explained. The inclusion criteria for normal subjects were as follows: (1) best-corrected visual acuity of 20/20 or better in both eyes, (2) no manifest tropia at distance or near fixation with simultaneous and alternate prism cover tests, (3) fusion at 0.33 m and 6 m with the Worth 4-dot test, and (4) no history or presence of ocular or systemic disease. A general ophthalmic evaluation was performed for all subjects, including best-corrected visual acuity, simultaneous and alternate prism cover test with fixation targets at 0.33 and 6 m, Worth 4-dot test at 0.33 and 6 m, the distance Randot stereotest, and the distance 3-D stereotest.

● DISTANCE RANDOT STEREOTEST: Distance stereoacuity was quantified using the random dot-based distance Randot stereotest (version 2; Stereo Optical Co, Inc.), because there is no available contour-based distance stereotest. This test is a Polaroid vectograph-based distance stereotest that is sensitive to disturbances or changes of distance binocularity.7–9 Subjects viewed the stereogram at a distance of 3 m while wearing polarizing glasses. The room illumination was recorded as 250 to 300 lux measured in front of the eye of the subject by the same illuminance meter mentioned previously. Each level of disparity consisted of two shapes. Testing began with coarse disparities that progressively became smaller as

● POLARIZED STEREOSCOPIC MONITOR:

The stimulus was presented on a 46-inch stereoscopic monitor (G460X; Pavonine Co., Inc., Incheon, Korea) with the background luminance of 250 cd/m2 (785 lux), resolution of 1920 ⫻ 1080 pixels, and contrast ratio of 1800:1 (Figure 1). Two images were displayed on the same monitor through different polarizing filters to present a stereoscopic stimulus. The subject wore glasses that also contained a pair of different polarizing filters. Each pair of filters allows light that has the same polarization state to pass through and blocks light that has an orthogonal polarization state, thus producing different stimuli for each eye. 1082

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

JUNE

2011

TABLE 1. Categorical Agreement between the Distance 3-Dimensional Stereotest and the Distance Randot Stereotest in Normal Adult Subjects

Distance Randot (arcsec)

800 to Nil 250 to 800 70 to 200 20 to 60

Distance 3-D (arcsec) 800 to Nil

250 to 800

70 to 200

20 to 60

0

2 8a 4

2 15 33a

0

arcsec ⫽ seconds of arc; 3-D ⫽ 3-dimensional. The number of subjects represented by each category is embedded in each category. a Identical results on distance 3-D stereotest and distance Randot stereotest.

FIGURE 2. Box-and-whisker plot showing measurable distance stereoacuity using the distance Randot stereotest (Distance Randot) and distance 3-dimensional stereotest (Distance 3-D) in 64 normal adult subjects. arcsec ⴝ seconds of arc; center lines ⴝ median; box tops and bottoms ⴝ quartiles; whiskers ⴝ extreme values.

follows: 400, 200, 100, and 60 arcsec. This is less compact than the distance 3-D stereotest and the near stereotests, such as the Randot Preschool stereotest with test disparities of 800, 400, 200, 100, 60, and 40 arcsec. If both responses were correct, testing proceeded to the next disparity until the subject could not identify either shape. The smallest disparity at which a subject identified both shapes correctly was recorded as the stereoacuity threshold. The tests were administered in the following order: distance 3-D stereotest (initial test), distance Randot stereotest, and distance 3-D stereotest (retest). ● STATISTICAL ANALYSES:

Continuous values were expressed as mean ⫾ standard deviation. Stereoacuity values were transformed to log arcsec for analysis. To evaluate the concordance of the distance 3-D and the distance Randot stereotests, stereoacuity results were grouped into 4 categories as fine (20 to 60 arcsec), moderate (70 to 200 arcsec), coarse (250 to 800 arcsec), and nil (⬎ 800 arcsec). The Wilcoxon matched pairs test was used to compare the stereoacuity test scores between distance 3-D and distance Randot stereotests. Because of the noncontinuous scale of stereotest scores, the Spearman correlation test was used to test the validity of the distance 3-D stereotest. Differences between test and retest scores were calculated for each subject using the results of the distance 3-D stereotest. The 95% limits of agreement were calculated.10 These values then were converted back to octave steps, which also can be described as doublings.11 Each doubling of the stereoacuity threshold corresponds to a change of 0.3 in the log transformed value; therefore, we divided the 95% limit of agreement values by 0.3 to calculate the number of VOL. 151, NO. 6

TABLE 2. Proportion of Subjects with a Categorical Stereoacuity as a Function of Test Distance Randot–Distance 3-D

Categorical difference ⱖ 1 level 35.9% (23/64) Categorical difference ⱖ 2 level 3.1% (2/64)

Distance 3-D Test-Retest

12.5% (8/64) 0% (0/64)

3-D ⫽ 3⫺dimensional.

octaves. Agreement between scores also was represented as Bland–Altman plots.10 Statistical analysis was performed using GraphPad Prism software (version 5.0; GraphPad Software, San Diego, California, USA). Results were interpreted as statistically significant when P values were less than .05.

RESULTS ● NORMATIVE DATA AND VALIDITY: Stereoacuity threshold scores were 43.3 ⫾ 32.9 arcsec (range, 180 to 20 arcsec) using the distance 3-D stereotest and 103.4 ⫾ 86.2 arcsec (range, 400 to 60 arcsec) using the distance Randot stereotest. Stereoacuity threshold scores were better with the distance 3-D stereotest, and this was statistically significant (P ⬍ .0001, Wilcoxon matched pairs test; Figure 2). A multivariate logistic regression analysis revealed that age, visual acuity, and eye alignment were not significantly correlated with the stereoacuity scores. Table 1 shows the concordance of distance 3-D and distance Randot stereotests in normal adults. The results for the 2 stereotests were identical in 64.1% and within 1 disparity level in 96.9% of normal adults. Scores obtained with the use of the 2 tests demonstrated a statistically

DISTANCE 3-D STEREOTEST

FOR

ADULT SUBJECTS

1083

was no overall tendency for the retest values to differ from the initial test values (Figure 3).

TABLE 3. Test–Retest Variability of the Distance 3Dimensional Stereotest in Normal Adult Subjects

● MAXIMUM HORIZONTAL DISPARITY: Distance 3-D Retest (arcsec)

800 to Nil 250 to 800 70 to 200 20 to 60

The maximum binocular horizontal disparity that can be fused with vergence was 1828 ⫾ 794 arcsec (range, 4000 to 500). Subjects could perceive the contour-based circle stereogram as a single circle from this level of binocular disparity. Beyond this level, the stereogram seen in each eye was visible as 2 separate circles.

Distance 3-D Test (arcsec) 800 to Nil

250 to 800

70 to 200

20 to 60

6a 1

7 50a

0 0

arcsec ⫽ seconds of arc; 3-D ⫽ 3 dimensional. The number of subjects represented by each category is embedded in each category. a Identical results on distance 3-D stereotest and retest.

DISCUSSION PREVIOUS STUDIES CONCERNING 3-D MONITOR-BASED STE-

reotests were conducted with a small number of children and did not measure the distance stereoacuity. In addition, there is lack of normative data, validity, and reliability of the 3-D monitor-based stereotests in previous studies.5,6,12,13 The normative data from the present study showed relatively high concordance with normative data obtained from the distance Randot stereotest and good test–retest reliability, which supports the validity and reliability of the distance 3-D stereotest used in this study. Normative data for the distance 3-D stereotest (43.3 ⫾ 32.9 arcsec) was better than that for the distance Randot stereotest (103.4 ⫾ 86.2 arcsec). The higher level of stereoacuity measured using the distance 3-D stereotest reflects the difference in stereogram— contour-based circles and random dot-based geometrical shapes in the distance 3-D and distance Randot stereotests, respectively. Fawcett reported discrepancies in stereoacuity between the contour-based and random dot-based stereotests at short distance.14 However, normative data for the distance 3-D stereotest was similar to the data obtained from the study of distance contour-based circles stereotest using the Baylor Video Acuity Tester (BVAT) and Binocular Vision System ([BVS], Medtronic Xomed Solan Ophthalmics, Jacksonville, Florida, USA) (50 ⫾ 32 arcsec).15 In our study of test–retest variability using the distance 3-D stereotest, the 95% limit of agreement was 1.55 octaves, and therefore a 2-octave change in threshold values is likely to represent real change in distance 3-D stereotest results (Figure 3). This value is smaller than that of the Frisby-Davis distance stereotest (1.99 octaves) and similar to that of the distance Randot stereotest (1.49 octaves).11 In addition, the test and retest scores of the distance 3-D stereotest were similar, indicating that minimal learning effect between examinations, which would have caused threshold values of retest scores to be better than the test scores. We also measured the maximum binocular disparity that could be fused with disparity vergence to determine the largest measurable disparity of true stereopsis in normal subjects using this 3-D stereotest. This could provide a reference range of true stereopsis in subjects with poor stereopsis using the distance 3-D stereotest. Erkelens re-

FIGURE 3. Bland–Altman plots showing test–retest variability represented as the distance 3-dimensional (3-D) stereotest in 64 normal adults. Upper and lower dotted lines show 95% limits of agreement with 95% confidence intervals. The half width of the 95% limit of agreement was 0.47 log arcsec (1.55 octaves) in the distance 3-D stereotest. arcsec ⴝ seconds of arc; SD ⴝ standard deviation; octave ⴝ 0.3 change of the log transformed value.

significant correlation (r ⫽ 0.324; P ⫽ .009, Spearman correlation test). In a categorical definition of stereopsis (fine, moderate, coarse, and nil), the normative data obtained from the distance 3-D stereotest exhibited good concordance with the data from the distance Randot stereotest, supporting the validity of the distance 3-D stereotest. Less than 5% of subjects showed a 2-grade or more difference between the 2 tests (Table 2). ● RELIABILITY: Test and retest results for the distance 3-D stereotest were identical in 87.5% and were within 1 disparity level in 100% of normal adults (Table 3). The 95% limit of agreement is represented on the Bland– Altman plots in Figure 3. The Bland-Altman plots suggest that the magnitude of the test–retest differences did not seem to be dependent on the level of stereoacuity. The half-width of the 95% limit of agreement was 0.47 log arcsec (1.55 octaves) in the distance 3-D stereotest. There

1084

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

JUNE

2011

ported that the fusional limits of retinal disparity of 4 different subjects between the 2 half-images of a large random-dot stereogram at 1.5 m ranges between 128 and 175 arcmin (7680 and 10500 arcsec).16 In our study, the maximum binocular horizontal disparity that can be fused with vergence (1828 ⫾ 794 arcsec) was much smaller than that of the previous report. The difference in maximum binocular disparities between the 2 studies is thought to be the result of the difference in retinal eccentricity, spatial frequency, brightness, and contrast between the 2 instruments for measuring stereopsis.17 The strengths of our study include the relatively large sample size, the development of a 3-D monitor contour-based distance stereotest for quantitative measurement of distance stereoacuity with broad ranges and compact intervals of disparities, as well as the comparison of results with those of a well-established stereotest. The 3-D monitor-based stereotest can measure stereoacuity precisely at smaller intervals, as well as a broad range up to the maximum limit of binocular fusion. Modification of the software to generate various shapes and diverse arrangements of stereograms could reduce the potential learning effects inherent in repetitive examinations. Furthermore, the brightness and resolution of the 3-D monitor used in our study (250 cd/m2, 1920 ⫻ 1080 pixels) were superior to monitors used in previous studies4,5 (100 cd/m2, 640 ⫻ 480 pixels; 58 cd/m2, 640 ⫻ 1024 pixels), which can measure distance stereoacuity. There are some limitations of the present study. First, the normative data of our study was limited to young adult subjects (age range, 23 to 39 years) only at a distance of 3 m. Previous studies have reported the results of distance stereotests for pediatric, adult, normal, and strabismic subjects.1,2,7,8 Adams and associates tested the FrisbyDavis distance stereotest at 2 different distances: 3 and 6 m.1 To overcome the limitation of our study, the distance 3-D stereotest should be performed in subjects of more diverse age ranges and at various distances. However, the stereoacuity thresholds can be affected by retinal eccentricity.18 Therefore, the circle size and distance between circles should be adjusted proportionally according to the test distance. Second, monocular cues are present in the large disparity levels of the contour-based circle stereotest, giving rise to false-positive results. Fawc-

ett14 reported that monocular cues are present in the large disparity level of contour-based stereotests like Titmus circle tests (circle, 1 to 4; ⱖ 140 arcsec) or Randot circle tests (circles, 1 to 2; ⱖ 200 arcsec). In this study, we performed the monocular cue test only at thresholds of 400 to 20 arcsec because all subjects were young adults with normal binocularity. However, it may be necessary to conduct the monocular cue test in all range of disparities if it is to be performed in patients with abnormal binocular function. Furthermore, to overcome the limitation of monocular cues in contour-based stereotests, the target stereogram should be placed at random positions or should be made based on random dots. Third, the direct illumination from the self-emitting 3-D monitor may influence pupil size, retinal illuminance, and contrast sensitivity, which are the factors that may affect the level of stereopsis.19,20 Fortunately, in this study, there was no significant difference in illumination of the room between the distance 3-D stereotest and the distance Randot stereotest (250 to 300 lux, both), although the brightness of the self-emitting 3-D monitor was 785 lux. Moreover, it is known that there is no significant reduction in stereoacuity for pupil sizes larger than 1.5 mm.20 Therefore, the impact of pupil size on stereoacuity may be considered to be clinically insignificant. But, the difference in retinal illuminance and contrast sensitivity between the distance 3-D stereotest and the distance Randot stereotest may influence the measurement of stereopsis. Therefore, this should be noted when comparing the results of stereotests with different monitor brightness or contrast ratios. In conclusion, data obtained with the distance 3-D stereotest showed good concordance with those of the distance Randot stereotest, supporting the validity of the distance 3-D stereotest. Moreover, the distance 3-D stereotest demonstrated relatively good test–retest reliability—similar to, or better than, conventional distance stereotests. The normative data set presented in this study can serve as a useful reference for quantitative assessment of binocular sensory status in clinical management. With its potential to test stereoacuity thresholds up to 5000 arcsec, the distance 3-D stereotest may be useful for defining the course of binocular vision abnormalities in subjects with poor stereopsis.

PUBLICATION OF THIS ARTICLE WAS SUPPORTED BY GRANT A092206 FROM THE KOREA HEALTHCARE TECHNOLOGY R&D Project, Ministry of Health and Welfare, Seoul, Republic of Korea. The authors indicate no financial conflicts of interest. Involved in Design of study (J.K., H.K.Y., Y.K., B.L., J.-M.H.); Conduct of study (J.K., H.K.Y., Y.K., J.-M.H.); Collection and management of data (J.K., H.K.Y., Y.K.); Analysis and interpretation of data (J.K., H.K.Y., J.-M.H.); Preparation of manuscript (J.K., H.K.Y., Y.K., J.-M.H.); and Review or approval of manuscript (B.L., J.-M.H.). All aspects of the research protocol were in compliance with the Declarations of Helsinki and were approved by the Institutional Review Board of Seoul National University Bundang Hospital (IRB no. B-0907/080 – 006). The authors thank Baek Lok Oh and Tae Min Ha for technical assistance and helpful comments.

REFERENCES 1. Adams WE, Hrisos S, Richardson S, Davis H, Frisby JP, Clarke MP. Frisby Davis distance stereoacuity values in visually normal children. Br J Ophthalmol 2005;89(11):1438 –1441.

VOL. 151, NO. 6

DISTANCE 3-D STEREOTEST

2. Wang J, Hatt SR, O’Connor AR, et al. Final version of the distance Randot stereotest: normative data, reliability, and validity. J AAPOS 2010;14(2):142–146. 3. Frisby JP, Clatworthy JL. Learning to see complex randomdot stereograms. Perception 1975;4(2):173–178. FOR

ADULT SUBJECTS

1085

12. Fujikado T, Hosohata J, Ohmi G, Tano Y. A clinical evaluation of stereopsis required to see 3-D images. Ergonomics 1996;39(11):1315–1320. 13. Kriegbaum-Stehberger B, Jiang X, Mojon DS. Performance of a new, 3D-monitor based random-dot stereotest for children under 4 years of age. Graefes Arch Clin Exp Ophthalmol 2008;246(1):1–7. 14. Fawcett SL. An evaluation of the agreement between contour-based circles and random dot-based near stereoacuity tests. J AAPOS 2005;9(6):572–578. 15. Yildirim C, Altinsoy HI, Yakut E. Distance stereoacuity norms for the mentor B-VAT II-SG video acuity tester in young children and young adults. J AAPOS 1998;2(1):26 –32. 16. Erkelens CJ. Fusional limits for a large random-dot stereogram. Vision Res 1988;28(2):345–353. 17. Howard IP, Rogers BJ. Binocular Vision and Stereopsis. New York: Oxford University Press, 1995:313–325. 18. Siderov J, Harwerth RS. Stereopsis, spatial frequency and retinal eccentricity. Vision Res 1995;35(16):2329 –2337. 19. Legge GE, Yuanchao G. Stereopsis and contrast. Vision Res 1989;29(8):989 –1004. 20. Lovasik JV, Szymkiw M. Effects of aniseikonia, anisometropia, accommodation, retinal illuminance, and pupil size on stereopsis. Invest Ophthalmol Vis Sci 1985;26(5):741–750.

4. Livingstone MS, Hubel DH. Stereopsis and positional acuity under dark adaptation. Vision Res 1994;34(6):799 – 802. 5. Fujikado T, Hosohata J, Ohmi G, et al. Use of dynamic and colored stereogram to measure stereopsis in strabismic patients. Jpn J Ophthalmol 1998;42(2):101–107. 6. Breyer A, Jiang X, Rütsche A, Mojon DS. A new 3D monitor-based random-dot stereotest for children. Invest Ophthalmol Vis Sci 2006;47(11):4842– 4846. 7. Adams WE, Leske DA, Hatt SR, et al. Improvement in distance stereoacuity following surgery for intermittent exotropia. J AAPOS 2008;12(2):141–144. 8. Holmes JM, Birch EE, Leske DA, Fu VL, Mohney BG. New tests of distance stereoacuity and their role in evaluating intermittent exotropia. Ophthalmology 2007;114(6):1215– 1220. 9. Leske DA, Birch EE, Holmes JM. Real depth vs Randot stereotests. Am J Ophthalmology 2006;142(4):699 –701. 10. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;327(8476):307–310. 11. Adams WE, Leske DA, Hatt SR, Holmes JM. Defining real change in measures of stereoacuity. Ophthalmology 2009; 116(2):281–285.

1086

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

JUNE

2011

Biosketch Jongshin Kim, MD, is a Chief Resident of Ophthalmology at Seoul National University Hospital (SNUH), Seoul, Korea. Dr Kim received his MD from Seoul National University College of Medicine in 2006. Dr Kim completed an internship and continues his residency in SNUH. He has the future plan of research in biomedical optics (eg, high resolution and deep penetration functional OCT) and robotic ophthalmic surgery as a PhD candidate at Korea Advanced Institute of Science and Technology (KAIST) from next year.

VOL. 151, NO. 6

DISTANCE 3-D STEREOTEST

FOR

ADULT SUBJECTS

1086.e1