P–096 Real-time ranking of single spermatozoa using artificial vision analysis of complex motility patterns during ICSI aimed at improving fertilization and blastocyst development

Abstract

Abstract Study question Can real-time artificial vision identify beneficial movement patterns of single spermatozoa in a cohort visualized in PVP during ICSI possibly enhancing fertilization and embryo development? Summary answer Artificial vision seems able to identify advantageous movement patterns of individual spermatozoa having a significant impact on both normal fertilization and blastocyst formation. What is known already Spermatozoa isolated from poor semen may reduce the quality of embryo development and blastocyst formation. Normal motility is dependent on general sperm morphology and characteristic movement of the flagellum enabling forward mobility. Spermatozoa roll as they swim. It is known that this rotational motion around their longitudinal axis promotes rheotaxis, which is a mechanism that allows the sperm to navigate to the site of fertilization. Therefore, it is possible that the characteristics of the rotational movement are related to sperm quality. Study design, size, duration Non-intervention study based on a cohort of 132 videos of in-vitro fertilization treatments with ICSI during which the sperm selection process was recorded up to sperm injection. The study was performed at one IVF center within a 6-month period. Injected spermatozoa and their corresponding oocytes were individually assessed from fertilization to blastocyst formation. Videos, where spermatozoa selected for injection could not be identified, were excluded. Relevant outcomes included normal fertilization (2PN), and blastocyst formation. Participants/materials, setting, methods Using a digitizer attached to an optical microscope (640 x 480 pixels), videos were recorded to include the sperm selection process, immobilization, and subsequent injection following standard ICSI protocols. Individual spermatozoa motility features were extracted using a proprietary computer-vision algorithm (SID, IVF 2.0 LTD). The rotational movements of spermatozoa were inferred by computing the variations of the mean intensity of the sperm in the video-sequence across time (MI). Main results and the role of chance Based on SID’s analysis, we found statistically significant differences between the median prominences of the MI of those injected spermatozoa that resulted in successful fertilization in comparison to those with failed fertilization (p-value=0.029, 28 negative fertilization, and 71 positive fertilization) using a one-tailed t-Student test with a significance level of 5%. We also found statistically significant differences between the median prominences of the MI of those spermatozoa that resulted in blastocysts in comparison with the spermatozoa-oocyte cohorts which didn’t reach the blastocyst stage (p-value 0.004, 51 with negative blastocyst formation and 48 with blastocyst formation). Limitations, reasons for caution The size of this database is modest, therefore a larger study with multiple clinics will be necessary to confirm the findings. Large prominence does not necessarily assurance successful fertilization or blastocyst formation since there may be other factors such as oocyte quality or the ICSI technique. Wider implications of the findings: Objective assessment of sperm rotational movement is difficult to quantify and to be objectively assessed during standard sperm selection. Real-time artificial vision tools such as SID could assist embryologists during the sperm selection process for ICSI. Trial registration number NA