Definition of the optimal criteria for identifying hyperactivated human spermatozoa at 25 Hz using in-vitro fertilization as a functional end-point.

Spermatozoa from fertile donors were used to select and validate criteria for the automated classification of hyperactivated (HA) motility using a computer-aided sperm analysis (CASA) system operating at 25 Hz. In the first phase of this analysis, 710 sperm trajectories were analysed and classified as forward progressive (FP), transition phase (TP) or HA. These tracks were then subjected to a CASA analysis and the various attributes of sperm movement defined for each category of motion. From this analysis, criteria were identified and subsequently validated that permitted the automated selection of FP, TP+HA and HA cells, with > 90% accuracy. To determine which attributes of movement best predicted the fertilizing potential of human spermatozoa, a detailed analysis of sperm motion was undertaken in patients undergoing in-vitro fertilization therapy. This analysis indicated that the single most important aspect of sperm movement was the incidence of HA motility (defined as curvilinear velocity > 90 microns/s, linearity < 20%, dancemean [amplitude of lateral head displacement/linearity x 100] > 45.8 microns) after 3 h of incubation. Using the latter criterion as the first incorporated variable, multiple regression equations were created that explained up to 50% of the variance in fertilization rates. None of the other patterns of motion (FP, TP, TP+HA) was correlated with fertilization rates, and none of the other published algorithms for identifying HA cells were as efficient as those described here.