After 18 Years of Trying, AI Helps Couple Conceive Their First Child

After nearly two decades of heartbreak and failed fertility treatments, one couple is finally expecting their first child—thanks to artificial intelligence.

The couple, who wish to remain anonymous, had spent 18 years undergoing multiple rounds of in vitro fertilization (IVF) at clinics around the world. Despite their efforts, they were unable to conceive due to a rare condition called azoospermia, in which no measurable sperm are found in the male partner’s semen.

In a typical semen sample, there are hundreds of millions of sperm. But in azoospermia, the sperm count is so low that none can be detected—even after hours of scanning under a microscope.

Eventually, the couple turned to the Columbia University Fertility Center, where they tried a revolutionary new method powered by AI: the STAR system—short for Sperm Tracking and Recovery.

Unlike conventional methods, STAR uses advanced imaging and machine learning to scan semen samples with extraordinary precision. In one case, expert lab technicians found no sperm after searching for two days. But the STAR system, using a microscope-connected camera and high-speed image analysis, found 44 sperm in under an hour.

For this couple, just three viable sperm were discovered. They were extracted and used to fertilize the wife’s eggs via IVF—leading to a successful pregnancy. Their baby is due in December.

“It took me two days to believe I was actually pregnant,” the wife said. “I still wake up wondering if it’s real.”

Dr. Zev Williams, who led the STAR system’s development, called it a breakthrough in male infertility treatment. “It’s like finding a needle across thousands of haystacks—gently and accurately, without damaging the sperm,” he said.

AI is already transforming fertility care by helping assess egg quality and embryo viability. Now, thanks to innovations like STAR, it may also offer hope to countless men diagnosed with conditions once considered untreatable.