One hundred fourteen semen samples from Chinese males were analyzed for routine semen parameters including the semen volume, sperm count, percentage motility, and percentage normal morphology. Of these 114 samples, 54 also had movement characteristics of seminal and swim-up sperm evaluated by the computer image analyzer system (Cellsoft; Cryo Resources Co., New York). All semen samples were subjected to the swim-up procedure to harvest the motile sperm before inseminations of human oocytes. Fertilization was considered to have occurred when at least one oocyte was observed with two or more pronuclei. Semen samples were classified as infertile (0% fertilization rate; N = 32) or fertile (greater than 0% fertilization rate; N = 82) before statistical analyses. There was a significant difference (P less than 0.005) in percentage normal morphology of seminal sperm between the fertile (mean +/- SE; 67.3 +/- 1.2%) and the infertile (59.3 +/- 2.2%) samples. The percentage normal morphology of seminal sperm correlated (r = 0.3049; P less than 0.002) with the fertilization rate and this parameter was selected by the multivariate stepwise discriminant analysis as the discriminator capable of predicting the fertilization rate with 57.9% accuracy. Statistical analyses of samples where sperm movement was also evaluated demonstrated that there was significant differences (P less than 0.01) between the fertile (N = 38) and the infertile (N = 16) samples in percentage normal morphology of seminal sperm (67.8 +/- 1.8% vs 56.2 +/- 2.6%) and curvilinear velocity of swim-up sperm (89.2 +/- 3.5 vs 68.2 +/- 7.2 microns/sec).(ABSTRACT TRUNCATED AT 250 WORDS)

No abstract

Evaluation of male fertility is based predominantly on results from semen analysis and determination of the sperm concentration is one of the main parameters of the analysis. The availability of a fully automated videomicrographic digital image analyser would offer both an objective and rapid method for determination of the sperm concentration. In the present study the sperm concentration in 327 semen samples was determined by haemocytometer according to the World Health Organization guidelines, and also by a computer-assisted digital image analyser system. Results were classified according to the routine procedure (haemocytometer) before statistical analyses. The computerized measurements caused a shift to the right in the frequency distribution of sperm concentration. Sperm concentrations were more often overestimated significantly (P less than 0.001) by the computerized measurements in semen samples with concentrations up to 80.0 x 10(6)/ml. This overestimation seemed to be caused by the presence of particles in seminal plasma that were recognized incorrectly as sperm by the computer program. The computerized digital image analyser gave an average sperm concentration of 2.2 +/- 0.6 x 10(6)/ml (mean +/- SEM) in 17 azoospermic semen samples while the routine procedure did not detect the presence of sperm cells. After removing the seminal plasma by washing and centrifugation with culture medium, and using the swim-up procedure to harvest motile sperm, the computerized measurements showed comparable results with the routine procedure for those sperm preparations (n = 44) with sperm concentrations greater than 5.0 x 10(6)/ml.(ABSTRACT TRUNCATED AT 250 WORDS)