篩檢全國學齡前幼童立體視覺－製作低成本亂點立體圖及與TN○立體感測試圖之比較

Abstract

Random-dot stereograms (RDS) do not have monocular cue in testing stereopsis. The absence of stereopsis as tested with RDS has higher correlation with diseases such as refractive error, amblyopia and straismus etc., as compared with line or contour stereograms. Commercially available RDSs have several drawbacks. Some do not consider the cross/uncross problem and thus require more convergence efforts to perceive. Some do not apply forced-choice and double-blind methods which is the basic requirements of good psychophysical tests. The other important concern is that the costs of these RDSs are too high to be used in nationwide screening. We designed on personal computer our own RDS. The algorithm is positional shift and logical operatons, such as AND, OR, XOR.. of computer graphics. With this algorithm, we can hide any figure into RDS. Photograghed directly from computer screen, we made four 300 sec-arc RDSs with geometric figures hidden in. The measure of stereopsis with our RDS and TNO met with each other. The subjective readibility and youngest testable age, about 3 year-old, of these tests also met with each other. Our RDS test has the advantages of (1) forced-choice, double-blind test method, (2) more fun and (3) low cost. The improvements from the old version of our RDSs include (1) uncrossed disparity for less convergence efforts, (2) higher dot density for higher brightness and (3) suitable picture size and color. The low-cost of this RDS test makes it especially suitable for nationwide screening of stereopsis in preschool children.