Automated scoring of multiprobe FISH in human spermatozoa

In the case of chromosomal aneuploidy in sperm wherein the incident rate is low and a large number of cells require scoring, automated methods that rely on computer software to segment and to count fluorescence signals are particularly necessary due to countless hours spent in reading slides and to the potential for interoperator differences. The purpose of this pilot experiment was to determine whether there were significant differences in the estimates of disomy frequency produced by automated versus manual scoring of signals for chromosome X, Y, and 18 in human sperm. The frequency of X18, Y18, XX18, YY18, and XY18 were determined in four separate normozoospermic samples. Slides were hybridized using a standard sperm FISH protocol for centromere-specific probes. Between 500 and 564, DAPI positive nuclei were captured from each sample and scored using the automated system, and the same slides were scored by a trained cytogeneticist, who was blind to the purpose of the study and the automated system results. None of the estimated frequencies was significantly different between manual and automated methods, regardless of whether individual slides or pooled results across all samples were compared. To our knowledge, this is the first report examining the validity of automated cell scoring in human spermatozoa. The results from this pilot exploration of sperm FISH suggest the comparability between automated and manual methods for estimating sex chromosome disomy and provide evidence that automated laser scanning of multiprobe sperm FISH should be explored further.