Columbia AI finds sperm in nearly 30% of men

Columbia University’s Star system located sperm in just under 30% of men who had previously been diagnosed with azoospermia, researchers reported. The finding allowed some patients to use their own sperm in in vitro fertilization and produced the first confirmed pregnancy from the method in 2025.

Star, short for Sperm Track and Recovery, was developed at the Columbia University Fertility Center. The system pushes semen or tissue samples through microfluidic chips-tiny devices with channels as thin as a human hair-while an imaging system captures roughly 300 frames per second. A machine learning algorithm analyzes the images in real time to identify sperm cells among debris and other cells. A robotic arm then isolates the identified sperm within milliseconds, avoiding centrifugation, a spinning process that can damage fragile cells.

Doctors can use the recovered sperm in IVF. In reported cases, Star identified about 40 times more sperm than manual searches by trained technicians and achieved 100% sensitivity, researchers reported.

The first pregnancy using sperm recovered by Star was confirmed in 2025. The couple, identified as Samuel and Penelope, had been trying to conceive for more than two years. Samuel had been diagnosed with Klinefelter syndrome, a genetic condition often associated with very low or absent sperm production. Penelope described the scan results and fetal movement with relief: “It’s starting to feel really real now, especially because I’m feeling movement… We had our anatomy scan, and everything is just looking so great.”

Zev Williams, director of the Columbia University Fertility Center, called the outcome “wonderful and special” and described the team’s reaction: “Everyone was just jumping up and down with joy.”

Azoospermia means no sperm are detected in a man’s ejaculate using conventional testing. It affects about 10% of infertile men and roughly 1% of all men. For some patients, sperm may be present only in extremely low numbers or confined to testicular tissue, making detection by standard lab techniques difficult.

Researchers and clinicians involved with the project noted larger clinical trials are required to validate Star’s performance across broader patient groups and to establish clinical protocols. They also identified practical and regulatory steps that must be completed before the system can be widely adopted in fertility clinics.