Australian researchers have developed an AI algorithm called SpermSearch that can accurately identify sperm in severely infertile men in a fraction of the time it takes scientists, providing hope to men who want to father biological children.
The most severe form of infertility, non-obstructive azoospermia (NOA), affects around 1% of men and around 5% of couples seeking fertility treatment.
NOA is a complete absence of sperm in the ejaculate due to abnormal sperm production and can be caused by genetics, hormonal imbalances, radiation and toxins and medications. To help them become fathers, men with NOA must undergo a procedure where a portion of their testes is removed and sperm manually extracted so it can be used to fertilize their partner’s eggs. It can take up to six hours for embryologists to tease apart the tissue sample to find and isolate sperm, which can be challenging given the presence of other cells and particles.
Researchers from the University of Technology Sydney (UTS) have developed an AI algorithm that does away with the manual labor required to find sperm in a tissue sample and does it quickly and accurately.
Called SpermSearch, the algorithm was trained using thousands of still microscopic images featuring sperm and high levels of other cells and debris, with only the sperm highlighted. This enabled the algorithm to recognize what a sperm looked like.
Researchers used healthy sperm and testicular tissue samples from seven men aged 36 to 55 diagnosed with NOA who’d already undergone sperm retrieval. The men had donated leftover tissue that was not used for the procedure.
The researchers then compared how the AI fared against an embryologist regarding the time it took to identify sperm and their accuracy. Results showed that the AI found more sperm overall, but some were only spotted by the embryologist and some by AI alone. The embryologist found 560 sperm, whereas the AI found 611. The algorithm identified sperm in less than 1000th of the time taken by the embryologist and was 5% more accurate than the embryologist.
The researchers note that this is a proof-of-concept study that requires testing in clinical trials. But, they say, it highlights the usefulness of using AI for sperm detection.
“For men diagnosed with non-obstructive azoospermia, ICSI with sperm retrieved from the testicles is the only realistic chance of having biological children,” said Carlos Calhaz-Jorge, the chair of the European Society of Human Reproduction and Embryology. “This is a preliminary study on the use of AI for finding healthy sperm in men experiencing this type of infertility. Finding healthy sperm under the microscope in fragments of testicular biopsies can be an arduous process. The prospect of using AI to make the process quicker and more accurate is very interesting. We need to see more research to build on these results.”
The study was presented at the European Society of Human Reproduction and Embryology Annual Meeting on 27 June 2023.