A team of researchers led by Dr. Steven Vasilescu from the University of Technology Sydney has developed an AI software called SpermSearch to assist in solving male infertility. The AI can rapidly identify viable sperm in samples taken from severely infertile men, doing so 1,000 times faster than a highly trained human observer. This technology is particularly beneficial for men with a condition called non-obstructive azoospermia (NOA), where they have no sperm in their ejaculate. Traditionally, embryologists manually search for healthy sperm in surgically removed testes tissue, a time-consuming and labor-intensive process.
SpermSearch significantly speeds up this process by analyzing photographs of tissue samples and instantly identifying viable sperm. The AI was trained using thousands of images of complex tissue samples to achieve this remarkable speed. While SpermSearch is not intended to replace embryologists, it acts as an assistive tool, saving critical time in identifying viable sperm for fertilization.
The ability to quickly find viable sperm is crucial in fertility treatments, as there is a limited time window for fertilization once eggs are collected. Male infertility has become a growing concern, with factors such as pollution, smoking, poor diets, lack of exercise, and stress contributing to a decline in sperm counts.
Other researchers are also exploring AI applications in addressing male infertility. For instance, another technique tracks the speed and movement of sperm tails to assess their health, while a different approach identifies DNA damage in individual sperm cells.
While these AI developments are promising, they are still in the proof-of-concept stage and will require further clinical trials before becoming commercially available. Nonetheless, such advancements offer hope for men facing infertility issues, potentially increasing their chances of fathering biological children. The University of Technology Sydney team is currently preparing for clinical trials, with the ultimate goal of achieving a live pregnancy using their AI technology.