Abstract:

In vitro fertilization-embryo transfer (IVF-ET), as a core method of assisted reproductive technology (ART), often faces the limitations of clinical outcomes due to the subjectivity of gamete and embryo quality assessment, laboratory procedural errors, and the complexity of developing individualized treatment plans. Artificial intelligence (AI), as a cutting-edge technological tool, can be deeply integrated into the entire IVF-ET treatment cycle, playing a key supporting role in ovarian stimulation, gamete and embryo assessment, laboratory management, and clinical decision-making. Its core mechanism lies in leveraging the powerful data processing capabilities of machine learning and deep learning algorithms to integrate vast amounts of clinical data, thereby accurately predicting the drug dosage of ovulation induction, and optimizing the follicle monitoring; utilizing the image recognition technology to enable the automated and objective grading of sperm motility and embryonic developmental potential, as well as non-invasive karyotype prediction; and combining the electronic witnessing system with robotic technology to significantly enhance the safety of sample tracking and the standardization of micromanipulation procedures. The comprehensive and multi-dimensional support provided by AI technology throughout the entire process effectively reduces human error, offering a critical technical support for improving the clinical pregnancy rates of IVF-ET and advancing ART toward a precision medicine model. Although existing AI technologies still face challenges, such as insufficient data standardization, their continuous optimization and clinical translation will strongly promote the standardized ART practices and technological innovation.

Key words: Artificial intelligence, Reproductive technology, assisted, Machine learning, Deep learning, Precision medicine, Embryo assessment, Gamete selection