Pregnant After 18 Years of Trying. Thanks to Artificial Intelligence

A child thanks to Artificial Intelligence, after 18 years of attempts. And all thanks to the Star system, used by the Columbia University Fertility Center. What is it capable of doing? We'll see shortly, but in the meantime let's tell the story of an American couple who have been chasing a child for almost twenty years, undergoing various cycles of assisted procreation all over the world. All cycles failed due to his infertility, caused by severe azoospermia. In his seminal fluid, to be clear, the spermatozoa that, in a normal condition are hundreds of millions, could not be found even after hours of meticulous research under the microscope. Not even one.
But the couple - who apparently did not intend to use sperm donation - decides to try a new approach at the Columbia University Fertility Center, which has developed the Star method, which uses artificial intelligence to help identify and collect those rare sperm hidden in seminal fluid that are difficult to find under a microscope by the human eye, no matter how careful. The researchers at the Center then examined the man's sperm with the artificial intelligence system, finding three individual sperm, which were collected and used to fertilize the woman's eggs, leading to the first pregnancy made possible by the Star method. The baby is expected in December.
AI in PMA laboratoriesArtificial Intelligence has entered the laboratories of reproductive medicine for a few years and it is no coincidence that entire sessions were dedicated to it at the last Eshre congress in Paris, which has just ended, focused on Reproductive Medicine and Embryology. AI helps laboratories evaluate the quality of oocytes, the health of embryos and could also be used in the field of male infertility, as in this case.
But let's get to the Star method, developed by the director of the Columbia University center, Zev Williams , and his colleagues, precisely to help identify the few spermatozoa present in samples from men with azoospermia. They themselves were struck by the results. "The Star system will make the difference, it's what you call a game changer. Consider that in a sample provided by a patient where highly trained technicians had not found even one spermatozoon, the Star system, in an hour, found 44".
When the sample is placed on a specially designed chip under the microscope, the Star (Sperm Tracking and Recovery) system connects to the microscope via a high-speed camera and uses high-precision imaging technology to examine the sample, taking more than 8 million images in less than an hour. The system then instantly isolates the detected sperm in a tiny droplet of culture medium, allowing embryologists who would never have been able to spot it with their own eyes to collect and preserve it.
“It’s like looking for a needle in a thousand haystacks, because this patient didn’t have the usual 200 or 300 million sperm, but two or three, literally. And the system, completing the search in less than an hour without using any means that could damage the sperm, found them,” Williams explained.
Male infertilityAccording to one estimate, in the United States, male infertility accounts for 40%, and up to 10% of infertile men are azoospermic. In Italy, the estimate of non-obstructive azoospermic men is about 2%. But discovering this is not easy, because these are healthy men, who have no problems in sexual relations and only realize they are healthy when they want to have a child and fail. "It is usually an unexpected and shocking diagnosis," Williams specifies. So far, attempts have been made to collect sperm directly from the testicle. "But it is an invasive, painful procedure (performed under anesthesia, ed. ) and can only be performed a couple of times."
Star is only available at Columbia University Fertility Center, now Williams and colleagues are planning to publish their results, sharing them with other centers. Using this method to find, isolate and freeze the found sperm currently costs a little less than $3,000.
What to do with current modelsStar is a new approach that will certainly help identify spermatozoa, but Artificial Intelligence has already entered as a protagonist in PMA laboratories around the world. "There are already models and algorithms used to analyze embryos at an early stage and accurately predict how healthy they are and how the implantation will go - explains Danilo Cimadomo , - associate professor at the Laboratory of Biology and Biotechnology of Reproduction at the University of Pavia and AI and data research manager for embryological research Ivirma - and other models that are able to evaluate the quality of oocytes before freezing to preserve fertility. Artificial Intelligence will certainly be increasingly important and Star is extremely interesting because when you have a testicular biopsy in the case of absence of spermatozoa in the ejaculate, it takes an operator who spends many hours screening under the microscope in search of one or more spermatozoa. This is a technique that without ICSI (single spermatozoon injected into a single oocyte, ed. ) was not possible at all and azoospermic patients could not reproduce at all. But it is a very laborious practice and requires many hours of operator time, so a tool of AI that screens the plate is extremely interesting because it takes less time than a man. It is an improvement because perhaps spermatozoa that had not been seen are identified, as in this case, and in any case it allows to save a lot of time. In my opinion it is very promising as a technique and will allow to standardize the procedures and support the clinical practice. However, it is not correct to say that the woman of that couple would not have gotten pregnant without the AI, perhaps they would have succeeded anyway, but certainly the AI facilitated and made the procedure possible”.
repubblica