A landmark fertility achievement, this artificial intelligence-driven intracytoplasmic sperm injection system marks a major step toward full automation in assisted reproductive technology.
First baby born via fully automated ICSI system | Image Credit: © MidoSemsem - © MidoSemsem - stock.adobe.com.
First baby born using fully automated system
The first baby was born using a fully automated, digitally controlled intracytoplasmic sperm injection (ICSI) system on April 10, 2025, as reported in Reproductive Biomedicine Online.1
Jacques Cohen, PhD, embryologist at Conceivable Life Sciences, led a multidisciplinary team of specialists alongside others to develop the automated system. It was developed to automate all 23 steps of ICSI, with either artificial intelligence (AI) or a remote operator independently performing the steps.
“With AI, the system autonomously selects sperm and precisely immobilizes its midsection with a laser ready for injection - executing this rapid, precise process with a level of accuracy beyond human capability,” said Mendizabal-Ruiz, PhD, lead engineer professor and an additional study head,
Studies have highlighted significant variation in ICSI performance between embryologists. According to Cohen, the automated system provides an opportunity to make the procedure more precise and efficient. Additionally, it may improve egg survival and optimize injection timing.
Case report: Patient and procedure details
The procedure occurred in a 40-year-old woman. After a failed in vitro fertilization (IVF) attempt, this patient was referred to Hope IVF Mexico for treatment with donor eggs. The study cycle included 5 eggs assigned for fertilization and 3 for control. Standard manual ICSI was provided for the control eggs.
A procedure duration of 9 minutes 56 seconds per egg was reported. While this time was slightly longer than the standard for manual ICSI, Mendizabal-Ruiz expects this time to be significantly reduced in future procedures.
Normal fertilization was accomplished for 4 of 5 eggs in the automated system and all control eggs. Through remote control in New York, 3700 km away from the site, a high-quality embryo successfully progressed to the blastocyst stage in culture. This blastocyst was transferred in the following cycle and led to the delivery of a healthy male infant.
Components and techniques used
Off-the-shelf components used to develop the system included an inverted microscope, a heated stage, a non-contact laser, a piezo actuator, a motorized stage, a motorized objective turret, oil and air microinjectors, and cameras.2 Lase sperm immobilization, laser-assisted ICSI, and piezo-ICSI were employed during the procedure.
The process involved 23 individual steps. Operators could perform these steps remotely through a computer-based interface, requiring no direct interaction with the micromanipulators or microscope. A standard bevelled microneedle is used for injection.
SiD (Mendizabal-Ruiz et al. 2022) was the AI incorporated into the system, able to determine the highest-ranked spermatozoon’s trajectory through software enhancements. Oocytes were identified by another in-house-developed AI to determine positions for holding pipette alignment.
Performance metrics and time comparisons
There were 115 total steps performed with the system, with a digital intervention rate of 49.6%. An average duration of 1 minute 39 seconds was reported for identifying and stabilizing an oocyte on a holding pipette, followed by an average of 3 minutes and 50 seconds for laser immobilization and pick-up.
Finally, the average time to deliver the sperm injection and release the oocyte was 4 minutes 38 seconds. In comparison, the manual procedure had an average time of 40 seconds for sperm selection, mechanical immobilization, and pick-up, and a total average per oocyte of 1 minute 22 seconds.
“While current limitations, such as operator involvement and procedure time, exist, these challenges are expected to diminish with further advancements in autonomy,” wrote investigators. “This work marks a sizable step towards automation of the IVF laboratory, which could significantly and positively impact the accessibility and efficiency of ART.”
References