A recent study has proposed the use of a new artificial intelligence (AI) tool that shows promising potential in quickly identifying sperm in infertile men. The aim of the research was to explore whether AI could speed up the treatment procedure for men who lack sperm in their semen.
At present, some patients undergo a process known as Intracytoplasmic Sperm Injection (ICSI). Embryologists inject a single live sperm directly into the center of a human egg.
However, this treatment can be time-consuming, taking up to six hours to locate and isolate sperm within the human tissue. To address this, a team of scientists from the University of Technology Sydney has been investigating methods to accelerate the process.
The results of the study were presented at the 39th annual meeting of the European Society of Human Reproduction and Embryology (ESHRE).
Training of AI tool ‘SpermSearch’
The team conducted their research at an IVF clinic in Sydney, dividing it into two phases that spanned over a period of five months. They utilized specialized AI software, which was installed on a computer, as their tool for investigation.
The researchers made the decision to train the AI tool, named SpermSearch, by exposing it to thousands of static microscope images.
These images consisted of sperm cells along with numerous other cells and debris. However, only the sperm cells were specifically highlighted for the AI’s focus.
Through this training process, the AI tool gradually learned to identify and distinguish sperm cells by analyzing the images.
Moreover, it developed its own evaluation system, allowing it to assess its performance and make necessary adjustments.
Dale Goss, the lead author of the study, and his team utilized both healthy sperm samples and samples of testicular tissue obtained from seven patients aged between 36 and 55.
What an honour to present our work on #AI to improve surgical sperm collection sperm searches. This has huge potential in routine clinical practice for non-obstructive azoospermia and we’re exciting to push this technology w/ @NeoGenixBio @StevenVasilescu @majidew @UTSFEIT pic.twitter.com/jusr46hG2b
— Dale Goss (@goss_dale) June 27, 2023
Some of these patients had already undergone surgical sperm retrieval at the clinic and were diagnosed with a condition called non-obstructive azoospermia (NOA).
‘AI took 1000th of the time taken by an embryologist’
In order to determine the most favorable outcomes, the researchers compared the time taken by both the AI tool and the embryologist to identify sperm, as well as their respective levels of accuracy.
The scientists discovered that the AI tool surpassed the embryologist in terms of total sperm detected. Specifically, the AI tool successfully located 611 sperm, while the embryologist found 560.
The study authors highlighted that the algorithm used by the AI tool exhibited exceptional efficiency, as it identified sperm within each segment of the droplet it examined in significantly less time than the embryologist—less than a thousandth of the time, to be precise.
See all the latest news from Greece and the world at Greekreporter.com. Contact our newsroom to report an update or send your story, photos and videos. Follow GR on Google News and subscribe here to our daily email!
