The evolution of preimplantation genetic testing for aneuploidy (PGT-A) techniques has been crucial in assisted reproductive technologies (ARTs), improving embryo selection and increasing success rates in in vitro fertilization (IVF) treatments. Techniques ranging from fluorescence in situ hybridization (FISH) to next-generation sequencing (NGS) have relied on cellular material extraction through biopsies of blastomeres at the cleavage stage on day three or from trophectoderm (TE) cells of the blastocyst. However, this has raised concerns about its potential impact on embryo development. As a result, there has been growing interest in developing non-invasive techniques for detecting aneuploidies, such as the analysis of blastocoel fluid (BF), spent culture medium (SCM), and artificial intelligence (AI) models. Non-invasive methods represent a promising advancement in PGT-A, offering the ability to detect aneuploidies without compromising embryo viability. This article reviews the evolution and principles of PGT-A, analyzing both traditional techniques and emerging non-invasive approaches, while highlighting the advantages and challenges associated with these methodologies. Furthermore, it explores the transformative potential of these innovations, which could optimize genetic screening and significantly improve clinical outcomes in the field of assisted reproduction.

To assess the performance of a novel microfluidic sperm selection device, SwimCount Harvester (SCH), compared with the Swim-up (SU) method, analyzing both sperm quality and the laboratory’s key performance indicators. Prospective, double-blind study using sibling oocytes. This study included 100 patients undergoing intracytoplasmic sperm injection. Both partners had normal karyotypes. Male participants provided only fresh ejaculated semen samples without total asthenozoospermia, globozoospermia, azoospermia, or cryptozoospermia. Female partners had at least 2 mature oocytes retrieved. Semen samples were analyzed to assess concentration, total progressive motile count, vitality, morphology, chromatin structure integrity, and deoxyribonucleic acid fragmentation index (DFI) before and after sperm selection. Subsequently, each sample was divided into two aliquots and processed using either the SCH device or the Swim-up technique. Retrieved oocytes from each patient were divided into two groups and microinjected with sperm from the corresponding preparation method. The injected oocytes were cultured in time-lapse incubators to collect morphokinetic data and apply artificial intelligence (AI)–based tools for assessing oocytes and embryos. Primary outcomes included fertilization rate, usable blastocyst rate, good-quality blastocyst rate, euploidy rate, embryo morphokinetics, and AI score to predict pregnancy for each embryo. Secondary outcomes encompassed oocyte quality (using AI tool) and sperm quality parameters. SwimCount Harvester yielded significantly higher sperm concentration, morphologically normal sperm, improved chromatin stability, and lower DFI. The usable blastocyst rates were higher in the SCH group (SCH, 40.5%, vs. SU, 34.5%), as well as the good-quality blastocyst rates (SCH, 30.8%, vs. SU, 23.4%). Time to blastocyst formation was shorter in the SCH group (SCH, 106.9 hours, vs. SU, 109.5 hours). Fertilization rates (SCH, 78.8%, vs. SU, 77.0%), euploidy rates (SCH, 40.4%, vs. SU, 37.2%), and AI scores to predict pregnancy (SCH, 5.7%, vs. SU, 4.6%) were comparable. Subgroup analyses revealed higher usable blastocyst and good-quality embryo rates in the SCH group among cases with a DFI of >20%, as well as a higher good-quality embryo rate in low-quality oocytes (AI score of ≤6) in the SCH group. SwimCount Harvester improved sperm quality compared with Swim-up. Furthermore, the SCH device increased the usable blastocyst rate and embryo quality rate and reduced time to blastocyst formation. Nevertheless, no significant differences were observed in the fertilization rates, euploidy rates, or AI scores.