Processes and Data Management of Computer-Aided Sperm Analysis in Human and Animal Spermatozoa

Abstract

Computer-Aided Sperm Analysis (CASA) has been revolutionized in the last two decades. It is now possible to measure a vast range of sperm characteristics such as sperm concentration, sperm motility parameters (including percentage of motility groupings, kinematics, subpopulations, mucous penetration, and hyperactivation), sperm morphology normality/types, sperm vitality, sperm DNA fragmentation, sperm acrosome reaction in many mammalian species but also in most vertebrate classes and some invertebrate phyla. While CASA methodology and analysis may be straightforward and performed within minutes, a major drawback is that during the analysis methodology variables (e.g., handling of semen sample, temperature control, correct optics, random field selection, and many more) are not observed/controlled in a manner that is consistent and repeatable. This paper emphasizes ways to control such variables. Advanced CASA systems can provide more than 100 quantitative parameters for sperm motility and morphology alone. Which of these sperm parameters are the most important in understanding sperm quality/physiology/potential to fertilize? What is the best way to analyze this large number of data points? In this regard, the paper recommends which parameters should be used (e.g., for motility subpopulations and hyperactivation) and which analysis techniques should be implemented, including aspects such as Bland and Altman plots, receiver-operating characteristic (ROC) curves, and various multivariate techniques including multivariate visualizations and principal component analysis.