Seven percent of the male population struggles with infertility. Now, artificial intelligence (AI) might be about to assist in finding a solution to the issue.
According to Dr. Steven Vasilescu, the artificial intelligence program that he and his team built can recognize sperm in samples collected from men who are severely infertile 1,000 times faster than a set of eyes that has been properly educated.
According to him, “It can highlight a potentially viable sperm before a human can even process what they’re looking at,” and this can happen before a human can even look at the sample.
Dr. Vasilescu is the creator of the medical startup NeoGenix Biosciences and also works as a biomedical engineer at the University of Technology Sydney (UTS), which is located in Australia.
The name of the program that he and his associates have built for it is SpermSearch.
It is intended to assist the 10% of infertile men who suffer from a disease known as non-obstructive azoospermia (NOA), which means that there are no sperm present in their ejaculate at all.
In situations like this, it is common practice to surgically remove a small piece of the testicles, which is then transported to a laboratory where an embryologist can manually search for viable sperm.
The tissue is carefully dissected before being analyzed using a microscope. It is possible to retrieve any viable sperm and then inject it into an egg, should this be necessary.
According to Dr. Vasilescu, this procedure can take a team of people up to six or seven hours, and there is a risk of both weariness and inaccurate results.
“What they see when they look down the microscope is just this complete mess – a starscape of cells,” he says. “When an embryologist looks down the microscope, what they see is just this complete mess.”
There is blood, and there is tissue. It’s possible that there are only ten sperm in the entire thing, but there might be millions of other cells. “It’s like looking for a needle in a haystack,” Dr. Vasilescu explains.
According to him, on the other hand, SpermSearch is able to locate any healthy sperm in a matter of seconds, and this is made possible by the fact that images of the samples are immediately uploaded into the computer.
In order to accomplish this level of speed, Dr. Vasilescu and his colleagues trained the AI to recognize sperm in these intricate tissue samples by presenting it with thousands of examples like these.
According to the findings of a test conducted by the UTS biomedical engineering team and detailed in a scientific report that was subsequently published, SpermSearch was 1,000 times faster than an expert embryologist.
On the other hand, the purpose of SpermSearch is not to take the role of embryologists but rather to serve as an auxiliary tool.
According to Dr. Sarah Martins da Silva, the urgency with which sperm must be located is of the utmost importance. “Time is critical,” says the clinical reader in reproductive medicine at the University of Dundee. “The clock is ticking.”
“There’s only a small time window for us to be able to do that if you’ve got somebody who’s had an egg collection and you’ve got eggs that need to be fertilized,” she said. “If you’ve got somebody that’s had an egg collection, you’ve got eggs that need to be fertilized.” It would be to everyone’s advantage to move the process along more quickly.
Infertility is still a rising problem, despite the fact that it is widely stated that sperm counts have dropped by half over the course of the last four decades.
According to some reports, the decline in male fertility can be attributed to a wide range of factors, including pollution and smoking as well as bad diets, insufficient amounts of exercise, and excessive amounts of stress.
Another academic who is striving to assist males who are dealing with infertility issues is Dr. Meurig Gallagher.
His novel method, which he developed while serving as an Assistant Professor in the Centre for Systems Modelling and Quantitative Biomedicine at the University of Birmingham, makes use of imaging software in order to monitor the velocity and activity of sperm tails.
He adds that observing the tail can provide information about the overall health of a sample. “Minor changes can tell us whether the sperm are responding to a biological cue, are about to die, or are under environmental stress,”
In the meantime, fertility company Examen, based in Belfast, makes use of a method known as single cell gel electrophoresis to analyze the level of DNA damage present in individual sperm.
Over the course of more than 20 years, Professor Sheena Lewis and her research group have been working to perfect the procedure.
However, Professor Lewis, who is both the chief executive of Examen and an emeritus professor in reproductive medicine at Queen’s University Belfast, believes that while recent breakthroughs in the use of AI are fascinating, the field of medicine progresses very slowly.
For instance, SpermSearch has reached the proof-of-concept stage of development after completing a very limited study with only seven participants.
“It doesn’t mean anything yet,” says Professor Lewis. “The amount of time that something takes to go from having a proof-of-concept stage to being accessible for purchase on a commercial scale is usually between two and five years.
“There is still a very long way to go. Additionally, it is geared toward the extremely limited population of males who suffer from NOA. Anything you are capable of doing is amazing, but you should know that it will never be popular.
When they got back to Sydney, Dr. Vasilescu told them that their therapy was the “last stop.”
“It can be the difference between fertilizing an egg – or just stopping treatment,” he explains. “It can be the difference between fertilizing an egg and…”
If we are able to make the embryologist more productive and accurate, they may discover sperm that they would not have discovered otherwise. That affords a man the opportunity to be the biological father of his own children.
The team at UTS is now prepared to begin testing their AI in clinical settings. According to Dr. Vasilescu, “an actual live pregnancy” is the next step that needs to be taken.
