The CellSoft computerized semen analysis system. I. Consistency of measurements and stability of results in relation to sample size analyzed.

The CellSoft TM computer-assisted digital image analysis system was evaluated for reproducibility of measurements and for the sample size needed to obtain stable results for seven parameters characterizing human spermatozoa. The parameters were sperm density, velocity, linearity, percent motile cells, maximum and mean amplitude of sperm head's lateral displacement (ALH max and ALH mean) and beat cross frequency. Thirty men were assigned to three equally large groups according to sperm density, and the groups were studied separately. Consistency of measurements was highly acceptable for all parameters (coefficient of variation ranging from 0.9% to 3.7%). The average minimum number of sample size needed for stable results varied between the groups and within each group dependent upon the parameter studied. Velocity and linearity seemed to require the lowest average cell number, while percentage motile cells and ALH max needed the highest sample size. Our current recommendation of sample sizes is outlined in detail and limitations and future utility of CellSoft is discussed.