This AI tool can find sperm in infertile men 1,000 times faster than a human

The new AI tool can identify sperm in seconds
The new AI tool can identify sperm in seconds Copyright Canva
By Luke Hurst
Share this article
Share this articleClose Button

A process that takes hours of careful work by an embryologist could be done in seconds by artificial intelligence (AI).


A new artificial intelligence (AI) tool may be capable of boosting the chances of severely infertile men producing children.

Currently those looking to become fathers rely on a time consuming process where embryologists manually search for viable sperm in samples taken from the patient.

While this process takes hours of careful work, the new AI tool can identify sperm in seconds. The study authors say this algorithm could bring hope to infertile men who want to father a biological child.

“This tool has the ability to give patients who have very little chance of fathering their own biological children an increased chance,” lead author Dale Goss, from the University of Technology Sydney, said.

“The algorithm improves antiquated approaches that have not been updated in decades. It will ensure the rapid identification of sperm in samples, which will not only increase the chance of a couple conceiving their own biological children, but also reduce stress on sperm and increase efficiency in the laboratory”.

From six hours to seconds

Roughly 1 per cent of men have the most severe form of infertility - known as non-obstructive azoospermia (NOA) - and they have no sperm in their semen.

The condition affects around 5 percent of couples seeking fertility treatment. They may, however, still have sperm in tiny quantities in the testes.

Currently, severely infertile men who want to find sperm have to undergo a procedure where a sample of their testes is removed.

This sample is then studied by embryologists, who manually identify and extract sperm from it, to then fertilise their partner’s eggs through Intracytoplasmic Sperm Injection (ICSI) treatment.

It can take up to six hours to find and isolate sperm in human tissue, which can undermine the embryologist’s ability to identify sperm because of mental and physical fatigue.

They have to carefully shred the tissue samples, releasing any sperm into a liquid in a petri dish. Then with a microscope, they search through the liquid, a tiny portion at a time, looking for sperm.

The process can be slowed down by contamination, and if the clinician misses the sperm, the patient has less chance to become a parent. The longer the process takes, the higher the chance is that the sperm won’t be viable.

The study, conducted by Australian experts, showed how the AI tool called SpermSearch can conduct this search in seconds, saving clinicians from the arduous process.

The clinicians can then decide whether the sperm is really present and if it is viable for ICSI. The results also show the algorithm is more accurate than experienced clinicians in identifying sperm.

The researchers trained SpermSearch first by showing it thousands of still microscope photographs, which featured sperm and high levels of other cells and debris, but only the sperm was highlighted.

The AI was able to then learn through image analysis what a sperm looked like.

Goss and his team used healthy sperm, and then samples of testicular tissue from seven patients aged from 36 to 55 years, all of whom had been diagnosed with NOA.

The algorithm was then tested against an embryologist whose precision was thought to be 100 per cent.


Comparing the time taken, the researchers found the AI could identify sperm in less than 1000th of the time taken by the clinician.

The AI also found more sperm overall, 611 compared to the embryologist’s 560.

The results of the study were presented at the annual meeting of the European Society of Human Reproduction and Embryology in Copenhagen, Denmark, on June 27.

In their conference presentation, the authors highlight that the study is based on a proof-of-concept test and that a clinical trial is required.

Share this article

You might also like