Stereopsis and perception of spatial depth in natural vision

Only the abstract of the article "Stereopsis and perception of spatial depth in natural vision" is available in English. The complete article can be read in German

Purpose: To investigate spatial vision under natural viewing conditions a new test is introduced, and it’s results are compared to stereoacuity measured by conventional stereopsis tests.

Material and Methods: The new test shows ten equal bottles at varying viewing distances of which subjects are required to name their sequence in a short period of time. The test was applied to 47 subjects with good eye health but varying stereoacuity. Both the test duration and the number of mistakes made were determined in one sequence where head movments were not allowed and in a second sequence
where head movements were allowed. Stereoacuity was deteremined as the median of three repeated measurements using a conventional stereopsis test. A Spearman rang correlation test was used to determine the correlation between test duration using the bottle test and stereoacuity as well as the number of mistakes using the bottle test and stereoacuity. To make assumptions for subjects with reproducible values for stereopsis only, the statistical analysis was repeated after removal of eight subjects from the sample whose stereoacuity exceeded 300arcsec.

Results: Statistical analysis shows a positive correlation between test duration for the bottle test and stereoacuity (r = 0,33; p = 0,02) as well as the number of mistakes for the bottle test and stereoacuity (r = 0,45; p = 0,001). Furthermore, it shows significant differences in both test duration (p = 0,02) and the number of mistakes (p = 0,04) when subjects with stereoacuity ≤ 100 arcsec and > 100 arcsec are compared. If all subjects with stereoacuity > 300 arcsec are excluded from the sample no significant differences were found between subjects with stereoacuity ≤ 100 arcsec and > 100 arcsec. In addition, the results show that head movements improve the performance for the bottle test.

Conclusion: The bottle test introduced in this study is suitable to asses spatial vision under natural viewing conditions. Subjects with good stereoacuity perform better than subjects with low stereoacuity. Performance deteriorates significantly if stereoacuity exceeds 300 arcsec. Hence, it is recommended to consider an increase of the commonly used stereoacuity threshold of 100 arcsec up to 300 arcsec.