Comparison of the measures of stereopsis using 4 different methods

Compartir 17/07/2015 Publicaciones Científicas

J.C. Ondategui-Parra1, R. Borras1, E. Peris2, Y. Sánchez2, S. Gómez-López1, J. Pujol1

1 Davalor Research Center (DRC) - Universitat Politécnica de Catalunya (UPC), Terrassa, Barcelona, Spain

2 University Vision Center (CUV) - Universitat Politécnica de Catalunya Terrassa, Barcelona, Spain.


To determine the agreement between 4 different clinical methods for measuring stereoacuity in near vision: Randot (R), Titmus-Wirt (TW), TNO and Frisby (F) in two populations groups: children and university students.


Measurements were performed in healthy young subjects and children. Noneof the subjects had strabismus, amblyopia, ocular disease or previous history of eyesurgery. Monocular best corrected visual acuity at far and near distances equal or betterthan 0.0 logMAR were required. The exams were performed in near vision (following thespecifications of each method) under controlled conditions of illumination (L≈450 lux).The lower limits were determined only by the threshold measurement of each test. Forthe analysis, the measures of stereopsis were transformed into a logarithmic scale.Finally, it was analyzed the agreement of the different stereoacuity results with thestandards clinical pass/fail criteria for each one.


The mean age of the 55 university students was 21.5±1.56 years (Range: 19to 23) and 9.2±0.45 (Range: 8 to 9) years for the 64 children. The range of stereoacuityresults in both groups were 20”-70” for R, 40”-70” for TW, 15”-480” for TNO and 20”-85”for F. In young subjects and children the differences between the values converted intoa log scale were not statistically significant for any test (p>0.05). Applying the Bland &Altman analysis, Confidence Interval (at 95%) more reduced were found comparing theR vs TW test and R vs F test, in both groups. In both groups the concordance was ≥95%in TW vs TNO tests and TW vs F tests. In addition, regarding children, it was obtained ahigh concordance TNO vs F tests. The lower agreement in the group of adults wasbetween the R vs F tests (74.5%) and for children between the R vs TNO tests (78%).(See table 1 and 2).


Studying the correlation between clinical tests of stereopsis is a complextask because (1) few tests allow the assessment of stereoacuity threshold due to thelower range of the results is imposed by the measurement tool; (2) the measurementscales are sometimes geometric progressions (TNO and F), or random measurementscales (R and TW), that difficult the analysis of the results.

The clinical cut-off criteria showed a high agreement in all cases. The pair of stereoacuitytests that presented a higher agreement was TW and TNO tests in both groups.

Comparison of the measures of stereopsis using 4 different methods


Todos los artículos

Juan Carlos Ondategui

Otros artículos