A Prospective, Parallel-group, Non-inferiority Study to Compare the Efficacy of an Automated Sperm Selection Method Versus Manual Sperm Selection for ICSI

Intracytoplasmic sperm injection (ICSI) is one of the standard clinical treatments for infertility. ICSI involves the injection of a single sperm into an oocyte with a sharp micropipette. Injecting a sperm with DNA fragmentation (i.e., physical breakage of the DNA double strands) into the oocyte deterministically lowers the IVF fertilization rate [1][2] and increases the miscarriage rate [3][4]. Patients with high sperm DNA fragmentation suffer from repeated IVF failures [8], causing heavy burdens on families and the healthcare system. Since the invention of ICSI in 1992, single sperm selection in ICSI has been made manually by embryologists, who select sperm by qualitatively choosing sperm with "good" motility and/or morphology based on their empirical experience. This involves significant subjectivity and inconsistency.

We have developed a robotic system to select sperm with low sperm DNA fragmentation. Automated sperm selection also eliminates the subjectivity and inconsistency in manual sperm selection. The system consists of a camera to acquire images of sperm and a software to analyze the images. The software automatically measures the 9 motility parameters (e.g., curvilinear speed, path linearity, etc.) and 11 morphology parameters (e.g., head ellipticity, midpiece width etc.). All these 20 parameters are defined by the WHO guidelines [9]. Embryologists select sperm by observing the same sperm characteristics as in the software criteria (e.g., speed etc.), but the software provides a more accurate and quantitative measure of sperm characteristics, thus ensuring the selected sperm have low DNA fragmentation.