With the recent shift to late marriage and late birth, the number of patients requiring infertility treatment is steadily increasing.The need for assisted reproductive technology such as in vitro fertilization is increasing year by year, and the number of assisted reproductive technology treatments conducted in Japan in 2016 reached the highest level in the world at about 45 cycles.
In "microinsemination", in which sperm are directly injected into an egg under a microscope, selecting good quality sperm is one of the important points to increase the fertilization rate.However, at present, there is no clear standard for discrimination, and it depends on the knowledge and experience of the assisted reproductive technology embryo culture specialist (hereinafter referred to as the embryo culture specialist) who carries out the work.In order to minimize the stress on the egg, it is necessary to quickly determine the sperm that seems to be the most suitable from among many sperms, which puts a heavy workload on the embryo cultivator and also cultivates the embryo. Differences in discrimination accuracy and working time between personnel are issues.
To solve these problems, the Jikei University School of Medicine announced that it has started joint research for the development of sperm discrimination assisting AI in collaboration with Olympus Corporation and Keiai Reproductive Medicine Clinic.
In this joint research, up to 1000 teacher data will be created from 1 cases, and AI will learn the criteria for discriminating sperm by comprehensively evaluating sperm head morphology and motility. The plan is to develop a sperm discrimination assist AI by December 2020 and aim to establish a microscope equipped with it.
As a result, good sperm can be recognized and displayed in real time to assist the embryo incubator's discrimination work, reducing the workload of microinsemination, as well as realizing homogenization of work and improving sorting technology. There is.