Australian researchers have developed an artificial intelligence algorithm called SpermSearch that can accurately identify the sperm of severely infertile men in a fraction of the time scientists need, offering hope to men who want to have children.
The most serious form of infertility is non-obstructive azoospermia (NOA), affecting about 1% of men and about 5% of couples seeking fertility treatment.
NOA is the complete absence of sperm in the ejaculate due to abnormal sperm production and can be caused by genetics, hormonal imbalances, radiation, toxins, and drugs. To help them become fathers, men with NOA must undergo surgery to remove part of their testicles and manually extract sperm so they can be used to fertilize their partners’ eggs. Embryologists can take up to six hours to comb through tissue samples to find and isolate sperm, which can be challenging given the presence of other cells and particles.
Researchers at the University of Technology Sydney (UTS) have developed an artificial intelligence algorithm that removes the manual labor required to find sperm in tissue samples and can do so quickly and accurately.
The algorithm, called SpermSearch, was trained using thousands of static microscopic images containing sperm and a host of other cells and debris, with only the sperm highlighted. This allows the algorithm to identify what the sperm looks like.
The researchers used healthy sperm and testicular tissue samples from seven men aged 36 to 55 who had been diagnosed with NOA and had undergone sperm retrieval. These individuals donated leftover tissue that was not used in surgery.
The researchers then compared the performance of the artificial intelligence and embryologists in terms of the time it took to identify sperm and their accuracy. The results showed that the AI found more sperm overall, but some only by the embryologists and some by the AI alone. The embryologists found 560 sperm, while the AI found 611. The algorithm identified sperm in less than one-thousandth the time of embryologists and was 5% more accurate than embryologists.
The researchers note that this is a proof-of-concept study that needs to be tested in clinical trials. But, they say, it highlights the usefulness of using artificial intelligence for sperm detection.
Carlos Calhaz-Jorge, president of the European Society for Human Reproduction and Embryology, said: “For men diagnosed with non-obstructive azoospermia, removal of sperm from the testes for ICSI is a good option. The only realistic chance to have a biological child.” “This is a pilot study using artificial intelligence to find healthy sperm in men with this type of infertility. Finding healthy sperm from testicular biopsy fragments under a microscope may be A daunting process. The prospect of using artificial intelligence to make the process faster and more accurate is very interesting. We need to do more research to build on these results.”
The study was published in Annual meeting of the European Society for Human Reproduction and Embryology June 27, 2023.