Optimization of a continuous real-time computerized semen analysis system for ram sperm motility assessment, and evaluation of four methods of semen preparation.

Sampling conditions that affect the biological validity of computer-assisted analysis of ram sperm motion were examined using a continuous real-time computerized semen analysis system (Hobson Sperm Tracker). Search radius (SR, 10 settings) and minimum track point (MTP, 10 settings) were varied factorially to evaluate their effects on the inclusion of sperm subpopulations within derived datasets. Low SR (< 12 microns) or high MTP values (> 26 frames) precluded measurements of rapidly moving cells, whereas high SR (> 17 microns) and low MTP settings (< 22 frames) led to erroneous tracking and poor data quality. Suitable settings for these set-up parameters were derived and tested for their biological consistency with four methods of preparing ram semen for computerized analysis. The preparation techniques tested were: centrifugation through sucrose-based Ficoll and Percoll media, a swim-up technique and simple dilution in Tyrode's media. The "selective' Percoll and swim-up methods generated sperm populations with significantly higher linear velocities and a lower tendency to deviate from linear trajectories than from either the Ficoll method or dilution technique. Deleterious effects of centrifugation were evident, particularly on sperm survival in vitro over several hours. It is concluded that computer-assisted semen analysis provides useful information about the behaviour of ram spermatozoa in vitro, but the measurement conditions must be defined carefully.