The couple, a 39-year-old man and his 37-year-old wife, had endured 19 years of infertility struggles, including multiple rounds of in vitro fertilization (IVF) and two invasive surgical procedures to extract sperm.

US couple has achieved pregnancy after nearly two decades of heartbreak, thanks to an innovative AI system that pinpointed elusive sperm cells invisible to the human eye.

The couple, a 39-year-old man and his 37-year-old wife, had endured 19 years of infertility struggles, including multiple rounds of in vitro fertilization (IVF) and two invasive surgical procedures to extract sperm.

What were the challenges faced by the couple?

Their challenges stemmed from the husband's condition of cryptozoospermia, a severe form of male infertility where semen samples appear normal but contain extremely rare or undetectable sperm under traditional microscopic examination. This issue contributes to up to 40 percent of infertility cases worldwide, often leading to repeated failures and emotional tolls.

How did the AI-based STAR scan method help?

Hope arrived at the Columbia University Fertility Center in New York, where researchers deployed the Sperm Tracking and Recovery (STAR) system, a cutting-edge AI-driven technology designed to revolutionise sperm selection for IVF. Led by Dr. Zev Williams, director of the center, the STAR method scans semen samples at unprecedented speeds, capturing and analysing millions of high-resolution microscopic images in real time to identify even the most scarce viable sperm amid cellular debris.

In this case, a routine 3.5-milliliter semen sample, deemed sperm-free by manual checks, was fed into the STAR system. Over approximately two hours, the AI processed a staggering 2.5 million images, detecting seven individual sperm cells: two motile (capable of movement) and five non-motile. Traditional methods, which rely on lab technicians manually scanning slides, would have missed these entirely, as they had in the couple's prior attempts.

The technology's ingenuity lies in its multi-layered process. High-powered imaging hardware first floods the sample with light, generating over eight million images per hour in standard runs. An advanced AI algorithm then sifts through the data, using machine learning trained on vast datasets to distinguish sperm from other particles with pinpoint accuracy. Once identified, a microfluidic chip, featuring tiny, hair-like channel, automatically isolates the sperm-rich portions of the sample.

A robotic arm then extracts the cells for immediate use in intracytoplasmic sperm injection (ICSI), where a single sperm is injected directly into an egg to create an embryo.

For the couple, the two motile sperm were selected and injected into two mature oocytes (egg cells). These developed into viable embryos, which were transferred to the wife's uterus on day three of the IVF cycle. Just 13 days later, a pregnancy test confirmed success. At the eight-week ultrasound milestone, doctors observed normal fetal development, including a strong heartbeat of 172 beats per minute, signaling a healthy progression. The patient has since transitioned to standard obstetric care.

"This is not just a win for one family, it's a game-changer for thousands facing similar barriers," said Dr. Williams in the study detailing the case, published today in the prestigious medical journal The Lancet. The research marks the first documented clinical pregnancy using STAR, highlighting its potential as a non-invasive, efficient alternative to surgical sperm retrieval, which carries risks like infection and scarring.