This paper presents a digital remote-controlled ICSI system, which innovatively integrates AI, laser, and piezoelectric technologies to achieve autonomous or digital control of 23-step full-process micromanipulation, and successfully delivered a live baby through remote control over 3,700 kilometers for the first time. This innovation verifies the feasibility of telemedicine in assisted reproduction, provides solutions for laboratory automation and standardization. Although the operational efficiency and full autonomy still need to be improved, it offers key clinical evidence and optimization directions for the digital transformation of assisted reproductive technology.
Research Background
In the field of assisted reproductive technology (ART), the operational accuracy of intracytoplasmic sperm injection (ICSI) has long relied on the professional skills of embryologists. However, issues such as operational variability caused by human factors, high-intensity training costs, and fatigue stress have persisted. Since the first clinical application of ICSI in 1992, its technical framework has not broken through the limitations of manual operation. Although early studies attempted automation of partial steps such as automatic sperm immobilization, clinical reports on full-process digital manipulation remain scarce. Based on this, this study aims to develop a remotely controllable fully automatic ICSI system, and explore its feasibility in human assisted reproduction through preclinical and clinical verification, providing a new path to solve the efficiency bottlenecks and standardization problems in traditional operations.
Research Methods: System Construction and Verification Design
(1) Technical Integration of Automated ICSI System
The remote-controlled system developed by the research team is centered on multidisciplinary technology integration. At the hardware level, it integrates Olympus IX81 inverted microscope, Tokai Hit heating stage, Lykos DTS non-contact laser, and Eppendorf PiezoXpert piezoelectric actuator. Submicron-level positioning control is achieved through the Prior Scientific motorized stage, combined with Arducam IMX477 camera for real-time image transmission. At the software level, it is equipped with the independently developed SiD artificial intelligence system, which uses deep learning algorithms for morphological segmentation and trajectory tracking of sperm head, midpiece, and tail. It can automatically screen high-quality sperm and guide the laser to focus on the middle part of the tail for double-pulse immobilization. The system can perform 23 micromanipulation steps in the full ICSI process, supporting AI autonomous control or remote digital intervention through a broadband network of ≥50Mbps, forming a closed-loop operation system.
(2) Preclinical and Clinical Experimental Design
Preclinical verification adopted a dual animal model system: in mouse experiments, 176 oocytes underwent piezo-ICSI after laser immobilization, with 173 using mechanical immobilization as controls, and fertilization rate, blastocyst formation rate, and live birth rate were recorded; in hamster models, 102 oocytes underwent ICSI by remote operators through a digital interface, with another 102 as manual injection controls, and oocyte survival rate at 2 hours was evaluated. The clinical study was approved by the Mexican CONBIOETICA registered IRB, enrolling a 40-year-old patient with primary infertility and AMH 0.1 ng/ml. 8 mature oocytes from a 23-year-old donor (5 in the automated group, 3 in the manual group) were used. Sperm were prepared by density gradient centrifugation and swim-up method. All fertilized eggs were cultured for 5 days before single embryo transfer.
Research Results: Outcomes of Animal Models and Clinical Applications
In preclinical verification, mouse model experiments showed that 176 oocytes undergoing piezo-ICSI after laser sperm immobilization had a fertilization rate of 98.8% and a blastocyst formation rate of 80.7%, which were not significantly different from the control group with 173 mechanically immobilized sperm (fertilization rate 99.4%, blastocyst rate 80.6%) (chi-squared test, P=0.57 and 0.93). The live birth rate after transplantation of 94 blastocysts in the laser group was 40.4%, which was not statistically different from the control group (47.8%) (P=0.31). In the hamster model, the 2-hour survival rate of 102 oocytes undergoing ICSI by remote operators through a digital interface was 94.1%, similar to 97.1% in the manual injection group (Fisher’s exact test, P=0.50).
In the clinical study, the 40-year-old patient used 8 mature oocytes from a 23-year-old donor (5 in the automated group, 3 in the manual group). The fertilization rate in the automated group was 80% (4/5), and 100% (3/3) in the manual group. Each group formed 2 usable blastocysts. After frozen-thawed blastocyst transfer in the automated group, the patient delivered a healthy male infant weighing 3.3kg via cesarean section at 38 weeks. This operation was completed remotely by a team in New York, 3,700 kilometers away, with only one on-site intervention needed due to oocyte position deviation.
B: The frozen-thawed blastocyst from remote ICSI was rewarmed and then re-expanded 4 hours later.
C: Six weeks after frozen-thawed blastocyst transfer, ultrasound showed a single pregnancy (including fetal heart).
Research Innovation: Technical Breakthroughs and Industry Value
The core innovation of this study lies in the first realization of digital closed-loop control over the full-process micromanipulation of ICSI: standardizing sperm selection through the SiD algorithm, combining laser-piezoelectric combined technology to replace traditional manual experience, breaking through the bottlenecks of geographical restrictions and human variability. The 49.6% autonomous operation rate of the system provides technical accessibility for resource-poor areas—for example, the Mexican laboratory connected to expert resources in New York through remote control, enabling successful pregnancy in an elderly patient with decreased ovarian reserve, demonstrating the potential of digital technology to optimize the allocation of medical resources. In addition, the established “AI execution + remote intervention” model redefines the role of embryologists, transforming them from traditional operators to system supervisors who adjust parameters such as laser energy through a digital interface, promoting ICSI towards standardization and intelligence.
Discussion and Conclusion: Practical Challenges and Future Directions
The current technology still has limitations such as long operation time and strong network dependence, and full autonomy needs to overcome technical bottlenecks such as dynamic adaptation to the microenvironment. Future optimizations will focus on lightweight AI algorithms, development of low-latency remote architectures, and automation of microfluidic sample pretreatment. The study confirms that the remote-controlled ICSI system achieves clinical live birth with an 80% fertilization rate, providing key evidence for the automated transformation of assisted reproductive laboratories. Its demonstrated standardization potential is expected to promote the accessibility and homogenization of global ART practices, opening a new chapter in digital assisted reproduction.
reference
Mendizabal-Ruiz G, Chavez-Badiola A, Hernández-Morales E, Valencia-Murillo R, Ocegueda-Hernández V, Costa-Borges N, Mestres E, Acacio M, Matia-Algué Q, Farías AF, Carreon DSM, Barragan C, Silvestri G, Martinez-Alvarado A, Olmedo LMC, Aguilar AV, Sánchez-González DJ, Murray A, Alikani M, Cohen J. A digitally controlled, remotely operated ICSI system: case report of the first live birth. Reprod Biomed Online. 2025 May;50(5):104943. doi: 10.1016/j.rbmo.2025.104943. Epub 2025 Apr 10. PMID: 40210512.