Columbia University Fertility Center has achieved the first confirmed pregnancy using an AI-guided sperm recovery method in a man with azoospermia, as detailed in The Lancet. The STAR system combined advanced imaging, microfluidics, and robotics to retrieve two viable sperm, enabling embryo creation after prior failed surgical and manual attempts.


Researchers at Columbia University Fertility Center have reported the first confirmed pregnancy achieved with an AI-guided sperm recovery method in a man with azoospermia, marking a notable milestone for severe male-factor infertility care, according to a case report described in The Lancet. The technique, called Sperm Tracking and Recovery (STAR), integrates advanced imaging, microfluidics, and robotics to find and retrieve extremely rare viable sperm from semen samples previously deemed sperm-free.
The reported case involved a couple attempting pregnancy for nearly two decades, including multiple IVF cycles, manual sperm searches, and two testicular surgeries before STAR was employed at Columbia. Using a 3.5 mL semen sample, STAR scanned approximately 2.5 million images in about two hours, identified two viable sperm, and enabled the creation of two embryos that led to a confirmed ongoing pregnancy.
Male factors account for about 40% of infertility among couples, and 10–15% of men with infertility have azoospermia, where ejaculate contains little or no sperm, significantly limiting options for biological parenthood. Conventional approaches include surgical sperm retrieval or prolonged manual searches after centrifugation, which can be invasive, time-consuming, costly, and risk damaging sperm, with inconsistent success. The AI-guided sperm recovery method offers a non-surgical, targeted alternative that may expand access to biological conception for patients historically told they had little chance.
The STAR method rapidly captures millions of images from a flowing semen sample and uses artificial intelligence to detect candidate sperm within cellular debris that obscures manual inspection under a microscope. A microfluidic chip with hair-like channels then isolates the specific portion of the sample containing the detected cell, and within milliseconds a robotic system gently removes the sperm for use in embryo creation or cryopreservation. Columbia’s team emphasizes that the pipeline was designed by specialists in advanced imaging, microfluidics, and reproductive endocrinology to reliably address each step of rare sperm identification and recovery.
“A semen sample can appear totally normal, but when you look under the microscope, you discover just a sea of cellular debris, with no sperm visible. Many couples with male-factor infertility are told they have little chance of having a biological child,” said Zev Williams, senior author and Director of the Columbia University Fertility Center, highlighting the diagnostic and therapeutic challenge STAR seeks to solve. Project lead Hemant Suryawanshi noted the multi-disciplinary design of the system to locate and isolate rare sperm in men with azoospermia, underscoring the method’s technical breadth.
While the first pregnancy using AI-guided sperm recovery is a proof-of-feasibility milestone, the current evidence rests on a single case report in The Lancet, requiring broader validation to assess generalizability, success rates, and clinical workflows across diverse patient populations. Columbia reports that larger clinical studies are underway to evaluate the efficacy of the STAR method in wider cohorts, which will be critical for determining patient selection, cost-effectiveness, and potential integration with IVF and ICSI pathways.
If supported by larger studies, AI-guided sperm recovery could reduce reliance on invasive testicular procedures and lengthy manual searches, offering a path to biological parenthood for some men with azoospermia who previously had limited options. As reproductive centers assess implementation, attention will turn to reproducibility, training, lab safety, and regulatory considerations for scaling the technology responsibly in clinical practice.
