Computer-assisted human semen analysis. Sampling errors and reproducibility.

Abstract

Videomicrographic computer-automated semen analysis systems allow quantitative description of sperm motility, velocity, progression, and head movement amplitude and frequency with unprecedented ease. The minimum number of spermatozoa needed for stable results, the variability of measurements and optimum methods of sampling the ejaculate were determined for one such system (Cell-Soft, CRYO Resources, New York, NY). Sampling a minimum of 225 spermatozoa yields stable measurements, and analyzing four microscope fields in triplicate provides data with the lowest coefficient of variation. The variability attributable to the instrument itself was acceptable for all measurements (6.2% to 15.4%) except mean amplitude of lateral head displacement. Limitations of these results and the potential utility of videomicrographic sperm movement analysis are discussed.