AI helping to create life, cherry picking of sperm cell . The famouse movie of Woody Allen, "Everything you always wanted to know about sex ( but were afraid to ask)" comes to my mind
Jorge, thanks for sharing this interesting article.
Men with azoospermia produce semen that appears normal but contains extremely few or no sperm—often fewer than three—making it nearly impossible to fertilize an egg naturally or through standard IVF. Typically, couples facing this struggle must use donor sperm or undergo invasive testicular surgery.
Developed over five years by Dr. Zev Williams and his team at Columbia University, STAR adapts techniques from astrophysics used to discover distant stars. It utilizes microfluidic chips, high-speed imaging, and AI algorithms: semen flows through a chip, is illuminated, and millions of images are captured per hour. The AI scans these images in real-time, identifying rare sperm amid cellular debris, and a robot isolates them.
In one trial, embryologists manually examined a sample for 48 hours and failed to detect any sperm. STAR, on the other hand, processed over eight million images in under an hour and found 44 sperm.
In March 2025, a couple who had endured about 18–19 years of unsuccessful IVF cycles finally conceived using sperm retrieved by STAR. Three sperm were successfully isolated and used to fertilize the woman’s eggs via IVF, culminating in a viable embryo transfer. The woman reported it took two days before the pregnancy felt “real,” and she remains around five months pregnant.
Advantages of STAR
Non-invasive: avoids surgical sperm extraction.
Fast and efficient: processes samples in ~1 hour.
Gentle handling: robotic extraction avoids damage from centrifugation or chemicals.
Cost-effective: sperm identification and banking costs under $3,000, much lower than full IVF cycles, which often exceed $12,400 to $15,000 in the U.S.
The Columbia team aims to expand STAR beyond azoospermia cases—possibly selecting optimal sperm, or even improving egg and embryo assessments. They hope to publish their data soon, enabling wider use across fertility centers.
Jorge, thanks for sharing this interesting article.
Men with azoospermia produce semen that appears normal but contains extremely few or no sperm—often fewer than three—making it nearly impossible to fertilize an egg naturally or through standard IVF. Typically, couples facing this struggle must use donor sperm or undergo invasive testicular surgery.
Developed over five years by Dr. Zev Williams and his team at Columbia University, STAR adapts techniques from astrophysics used to discover distant stars. It utilizes microfluidic chips, high-speed imaging, and AI algorithms: semen flows through a chip, is illuminated, and millions of images are captured per hour. The AI scans these images in real-time, identifying rare sperm amid cellular debris, and a robot isolates them.
In one trial, embryologists manually examined a sample for 48 hours and failed to detect any sperm. STAR, on the other hand, processed over eight million images in under an hour and found 44 sperm.
In March 2025, a couple who had endured about 18–19 years of unsuccessful IVF cycles finally conceived using sperm retrieved by STAR. Three sperm were successfully isolated and used to fertilize the woman’s eggs via IVF, culminating in a viable embryo transfer. The woman reported it took two days before the pregnancy felt “real,” and she remains around five months pregnant.
Advantages of STAR
Non-invasive: avoids surgical sperm extraction.
Fast and efficient: processes samples in ~1 hour.
Gentle handling: robotic extraction avoids damage from centrifugation or chemicals.
Cost-effective: sperm identification and banking costs under $3,000, much lower than full IVF cycles, which often exceed $12,400 to $15,000 in the U.S.
The Columbia team aims to expand STAR beyond azoospermia cases—possibly selecting optimal sperm, or even improving egg and embryo assessments. They hope to publish their data soon, enabling wider use across fertility centers.