New stereoacuity test using a 3-dimensional display system in children.

Abstract

The previously developed 3-dimensional (3D) display stereoacuity tests were validated only at distance. We developed a new stereoacuity test using a 3D display that works both at near and distance and evaluated its validity in children with and without strabismus. Sixty children (age range, 6 to 18 years) with variable ranges of stereoacuity were included. Side-by-side randot images of 4 different simple objects (star, circle, rectangle, and triangle) with a wide range of crossed horizontal disparities (3000 to 20 arcsec) were randomly displayed on a 3D monitor with MATLAB (Matworks, Inc., Natick, MA, USA) and were presented to subjects wearing shutter glasses at 0.5 m and 3 m. The 3D image was located in front of (conventional) or behind (proposed) the background image on the 3D monitor. The results with the new 3D stereotest (conventional and proposed) were compared with those of the near and distance Randot stereotests. At near, the Bland-Altman plots of the conventional and proposed 3D stereotest did not show significant difference, both of which were poorer than the Randot test. At distance, the results of the proposed 3D stereotest were similar to the Randot test, but the conventional 3D stereotest results were better than those of the other two tests. The results of the proposed 3D stereotest and Randot stereotest were identical in 83.3% at near and 88.3% at distance. More than 95% of subjects showed concordance within 2 grades between the 2 tests at both near and distance. In conclusion, the newly proposed 3D stereotest shows good concordance with the Randot stereotests in children with and without strabismus.