Particle.news

Download on the App Store

AI-Driven STAR System at Columbia University Yields First IVF Pregnancy

Developed over five years at Columbia University, the STAR system has located rare sperm in azoospermic samples to achieve a pregnancy now five months along; peer-reviewed findings are under review with clinic rollout in preparation.

As artificial intelligence shapes fertility care, it’s now helping doctors at Columbia University Fertility Center find hidden sperm.
Image
An embryologist works with advanced imaging equipment while an embryo appears on the computer screen, symbolizing the high-precision fertility techniques used in the STAR system—an AI-powered method that helped detect hidden sperm in azoospermic men, leading to successful pregnancies after years of infertility.
Image

Overview

  • STAR leverages AI, high-speed imaging and a microfluidic chip to capture over 8 million images per hour for detecting and isolating viable sperm from azoospermic samples.
  • In lab tests, STAR found 44 sperm in an hour in samples where embryologists detected none after two days, demonstrating its sensitivity.
  • The first clinical case used three sperm isolated by STAR in March 2025 to fertilize an egg via IVF, leading to a pregnancy that is now five months along.
  • STAR offers a non-invasive and lower-cost alternative to surgical sperm extraction, costing under $3,000 compared to $12,000–$30,000 typical IVF expenses.
  • Currently exclusive to Columbia University Fertility Center, the STAR team is preparing a peer-reviewed publication and outreach efforts to expand access to other fertility clinics.