Columbia's STAR AI Finds Hidden Sperm in 10% of Infertile Men
Updated
Updated · The New York Times · Aug 11
Columbia's STAR AI Finds Hidden Sperm in 10% of Infertile Men
2 articles · Updated · The New York Times · Aug 11
Summary
Dr. Zev Williams and Columbia’s fertility team developed STAR, an AI system that detects rare “hidden” sperm in men long thought to have none.
Up to 40% of infertility cases worldwide involve male factors, and azoospermia affects about 10% of infertile men, making missed sperm especially consequential.
STAR—short for Sperm Tracking and Recovery—uses machine-learning methods likened to astronomy tools that spot faint objects invisible to the human eye under a microscope.
For men with nonobstructive infertility, where sperm production itself is impaired, finding even a few viable sperm could reopen a path to biological fatherhood.