Columbia researchers report first successful pregnancy using AI-guided sperm retrieval method

Researchers at the Columbia University Fertility Center have reported the first successful pregnancy using an AI-guided method they developed to retrieve sperm from men with azoospermia, in which ejaculation contains little or no sperm.

The case is described in an investigative letter published in The Lancet.

Male factors are responsible for about 40% of couples with infertility. Of these, approximately 10-15% of men with infertility have azoospermia.

A sperm sample may look completely normal, but when you look under the microscope you just discover a sea of ​​cellular debris, with no sperm visible. Many couples with male factor infertility say they have little chance of having a biological child.”

Zev Williams, senior author of the paper and Director of the Columbia University Fertility Center

Men with azoospermia can undergo surgery to extract sperm from the testicles, but the procedure is often unsuccessful and can cause vascular problems, inflammation or a temporary decrease in testosterone levels.

Some specialized laboratories employ technicians to manually inspect sperm samples—a time-consuming and expensive process—after they have been processed by centrifugation or other agents that can damage the sperm.

“The field is really challenged to find a better way to identify and retrieve viable sperm in men with extremely low sperm counts,” says Williams.

A STAR is born

Williams assembled a team of researchers and clinicians to develop a new method that combines a variety of technologies to identify and retrieve rare sperm from men with azoospermia.

“Our team included experts in advanced imaging, microfluidic and reproductive endocrinology techniques to address every single step required to find and isolate rare sperm,” says Hemant Suryawanshi, assistant professor of reproductive sciences at Columbia University’s Vagelos College of Physicians and Surgeons and project leader.

Introduced earlier this year, the STAR (Sperm Tracking and Recovery) method uses high-powered imaging technology to scan a sperm sample from men with azoospermia, taking over 8 million images in less than an hour. Artificial intelligence is used to identify the sperm in the sample, and a microfluidic chip with tiny, hair-like channels isolates the part of the sperm sample that contains the sperm. Within milliseconds, a robot gently removes the sperm so it can be used to create an embryo or frozen and stored for later use.

First successful pregnancy using STAR

STAR was tested on a patient who had been trying to start a family for nearly 20 years, including multiple cycles of IVF at other centers, several manual sperm searches and two surgeries to extract sperm.

The patient provided a 3.5 mL semen sample. In about two hours, STAR scanned 2.5 million images, locating 2 viable sperm, which were then used to create two embryos and start a pregnancy.

The findings, although based on a single case, show the feasibility of this technology to overcome longstanding barriers to help men with azoospermia.

“You only need one healthy sperm to create an embryo,” says Williams.

Larger clinical studies are underway to evaluate the effectiveness of STAR in broader patient populations.