An algorithm for selection of ICSI sperm by computer vision firstly confirms that a sperm cell’s motility and morphology reflect its DNA fragmentation, and further suggests that selecting sperm quantitatively may mitigate the risk of DNA fragmentation.
The selection of sperm cells for ICSI remains an empirical procedure performed by embryologists and dependent largely on morphology and motility. Testing for DNA fragmentation is only possible on samples, which, when selected as individual cells for ICSI, may or may not meet the embryologist’s (and the WHO’s) criteria. Sperm selection for ICSI thus remains a qualitative and not quantitative exercise, and, according to a new report, ‘involves subjectivity and inconsistency’. That report comes from a new study of whether a sperm’s motility and morphology parameters reflect its DNA integrity, and whether quantitative (and not just qualitative) criteria can be applied in the selection of a single sperm with high DNA integrity for ICSI.(1)
The testing of sperm samples for DNA fragmentation remains somewhat controversial, although, as the authors of this study note, there is evidence of an inverse association between DNA fragmentation and fertilisation and blastocyst formation rates. The relationship between a single sperm’s motility and morphology and its DNA fragmentation level is thus ‘assumed’, they add, reliant on findings from semen samples but not at the single-cell level.
This new study, from the University of Toronto, tested a quantitative approach on a sample of 440 sperm cells randomly taken from 17 men. Motility and morphology parameters, which were defined for reference (with some adjustments) from the WHO’s criteria, were quantitatively measured with custom-developed computer vision algorithms, with images of live sperm captured at 30 frames per second. After the motility and morphology parameters were measured, the sperm was transferred for DNA fragmentation testing by the comet assay.
The sperm selection software selected 60 sperm cells as ‘normal’ – that is, every motility and morphology parameter satisfied the algorithm’s criteria, whereas the other 380 sperms had at least one abnormality. Moreover, the 60 selected normal sperms had significantly lower DNA fragmentation levels than the non-selected abnormal sperm. This testing was able to establish that the selection of single sperm according to these WHO qualitative criteria for motility and morphology parameters does achieve a higher DNA integrity than that of the overall sperm sample. Moreover, when the quantitative criteria were integrated into a software programme for sperm selection, the software outperformed the embryologists in blind tests by selecting sperm with the highest DNA integrity. The authors thus propose that the technique ‘holds the potential to mitigate the risk factor of DNA fragmentation’ in sperm selection for ICSI, while adding that the non-invasive technique ‘is presently capable of measuring 8 of the 11 WHO-defined morphology parameters’.
And they thus conclude: ‘The adapted WHO quantitative criteria help reduce the subjectivity of manual selection, enable automated sperm selection for ICSI, and hold the potential to standardize sperm selection in clinical ICSI.’
Meanwhile, another report reflects the extent to which computer technology is helping explain the viability of sperm cells in natural sperm selection and successful sperm migration. A group of andrology-lab scientists in Germany have reported their development of a ‘spatio-temporal’ computer simulation model of a mammalian (bovine) female genital tract in which sperm cells treated as agents with ‘concrete properties’ and specific rules for motion and interaction ‘swam’ against a fluid stream.(2) The simulation results, report the authors, demonstrated a close match with the observed timing and number of sperm actually reaching the entry of the oviducts, when compared with published data in cattle.
Among findings from the experiments were that physical sperm characteristics, such as velocity and directional stability, are essential for success. In addition, the ability to swim against the mucus flow of cervical secretions and align with epithelial walls of the genital tract ‘turned out to have a tremendous impact’ on the chances of a successful transit to the oviduct.
1. Zhang Z, Dai C, Shan G, et al. Quantitative selection of single human sperm with high DNA integrity for intracytoplasmic sperm injection. Fertil Steril 2021; doi.org/10.1016/j.fertnstert.2021.06.016
2. Diemer J, Hahn J, Goldenbogen B, et al. Sperm migration in the genital tract—In silico experiments identify key factors for reproductive success. PLOS Computational Biol 2021; doi.org/10.1371/journal.pcbi.1009109