Columbia University reports first pregnancy using AI sperm recovery

Columbia University Fertility Center has reported the first successful pregnancy using an innovative AI-guided sperm recovery technique called STAR (Sperm Tracking and Recovery). This non-invasive method was developed to address male-factor infertility in men with azoospermia, a condition characterized by little or no sperm and affecting 10-15% of infertile men. Traditional sperm retrieval methods, including surgical extraction and manual sample inspection, often fail or carry risks such as inflammation and hormonal issues. The STAR method combines advanced imaging, artificial intelligence, microfluidics, and robotics to scan millions of images of a semen sample, identify rare viable sperm cells, and gently isolate them for use in fertility treatments. In the reported case, a patient who had struggled with infertility for nearly 20 years and undergone multiple unsuccessful IVF cycles and surgical sperm retrievals provided a semen sample that the STAR system scanned, analyzing 2.5 million images over two hours. The AI successfully located two viable sperm cells, which were used to create embryos