AI Helps Locate Hidden Sperm in Men Once Deemed Azoospermic

Male factors contribute to up to 40 percent of all infertility cases worldwide. Among men seeking treatment, roughly 10 percent are diagnosed with azoospermia—an absence or near-absence of sperm in the ejaculate. Azoospermia can be obstructive, caused by a fixable blockage, or non-obstructive, in which the biological machinery that generates sperm is impaired, dramatically reducing the likelihood of fathering children.

New York Times

Dr. Zev Williams, director of the Columbia University Fertility Center, has led a team that created the STAR (Sperm Tracking and Recovery) system. Drawing an analogy to astronomers’ use of machine-learning to discover faint celestial objects, the STAR platform applies artificial-intelligence algorithms to microscope images of semen samples. The goal is to detect individual sperm cells that human observers often miss, especially in men whose counts are extremely low and who have previously been labeled azoospermic.

New York Times

The STAR workflow begins with a standard semen sample that is imaged at high magnification. AI models, trained on thousands of annotated images, scan each frame to flag motile or morphologically normal sperm. The software then highlights these “hidden” cells for embryologists to retrieve, enabling the use of sperm that would otherwise be discarded. This approach supplements, rather than replaces, conventional manual microscopy.

New York Times

If the AI-assisted detection proves reliable across larger patient cohorts, men with non-obstructive azoospermia could gain access to viable sperm for intracytoplasmic sperm injection (ICSI) or other assisted-reproduction techniques. Dr. Williams suggests that the technology may transform a diagnosis once considered “heartbreaking, devastating and surprising” into a more hopeful outlook for prospective fathers.

New York Times

• “That is really one of the most heartbreaking, devastating and surprising diagnoses you can give to patients,” — Dr. Williams, director of the Columbia University Fertility Center

New York Times