Columbia fertility clinic uses AI system to find rare sperm

In short: Columbia University Fertility Center is using an AI-based system called STAR to find and collect extremely rare sperm that standard lab methods can miss.

What happened

Columbia University Fertility Center has introduced a technique called STAR that searches for sperm in semen samples from men with azoospermia, which means no sperm is found on typical tests, or with extremely low sperm counts.

STAR puts the sample on a small chip and scans it quickly. The system takes millions of images in under an hour, then uses AI, which is software that spots patterns, to flag cells that look like sperm. A robotic tool then retrieves the sperm it finds.

The goal is to recover sperm that are so rare that a person looking through a microscope might not find them, or might damage them during handling. Reporting in 2025 said the approach had already led to a pregnancy for at least one couple after years of infertility, but the method was described as newly available and centered at Columbia.

Why it matters

For some couples, the biggest barrier is simply finding a usable sperm cell. If STAR can reliably find and pick up these rare cells, doctors may be able to use them for ICSI, where a single sperm is injected into an egg (like placing one seed directly into a hole instead of scattering many), or freeze them for later. It is important to note that ICSI itself is not new, the newer part is the AI-assisted search and fast retrieval.

Source: NYTimes