A research group consisting of Professor Tomoki Hamagami of Yokohama National University Institute of Engineering and Associate Professor Yasushi Yumura of Yokohama City University Citizens' Medical Center is supporting sperm evaluation work performed in male fertility treatment by artificial intelligence (AI). Developed a system to do.
According to Yokohama National University and others, the research group will collect sperm evaluation data and develop a system to train AI so that large-scale sperm data can be collected with the cooperation of embryo culture specialists working in clinics around the world. bottom.
Furthermore, we have developed a method for estimating the distribution of sperm evaluation with high accuracy, rather than simply evaluating it.This makes it possible to predict and present the evaluation results of a large number of embryo culture specialists.The error of the distribution system is 0.00035 (mean absolute value error 1.43%) for the training data and 0.025 (11.08%) for the verification data, confirming that the distribution can be estimated with high accuracy.
Intracytoplasmic sperm injection is a treatment method for patients with low sperm count in male infertility, but this method requires high cell discrimination ability to find promising sperm in a limited short time, which is a burden on embryo culture specialists. Is getting bigger.In order to increase the success rate of microinsemination, it has been required to develop a system that assists in the search and evaluation of sperm.
The research group believes that the development of this system will enhance the testing of male infertility, improve the fertility rate during infertility treatment, reduce the cost burden on patients, and pass on skilled embryo culture techniques.
reference:[Yokohama National University] Reproductive medical support AI cloud system developed (PDF)