Particle.news
Download on the App Store

AI-Guided STAR Recovers Rare Sperm, Enabling First Pregnancy Reported in The Lancet

Larger clinical studies are underway to evaluate the method's performance across broader patient groups.

Overview

  • Columbia researchers combined high-resolution imaging, artificial intelligence, microfluidics, and robotics to detect and retrieve individual sperm from debris-heavy samples.
  • In the reported case, the system analyzed a 3.5 mL semen sample for about two hours and identified two viable sperm that were used to create two embryos, initiating a pregnancy.
  • The platform can capture millions of images per hour to flag candidates for gentle robotic retrieval via a microfluidic chip.
  • The approach is designed to reduce reliance on testicular surgery and time-consuming manual searches that can fail or damage sperm.
  • Male factors account for roughly 40% of infertility, and 10–15% of infertile men have azoospermia, highlighting the potential clinical need the technology seeks to address.