Advancements in AI Enhance Embryo Selection and Hormonal Treatment Responses
The advent of artificial intelligence (AI) is bringing forth new hopes in the field of assisted reproduction, from identifying embryos with higher potential to improving responses to hormonal treatments and even shortening access times. Nearly half a century after the birth of the first child conceived through in-vitro fertilization (IVF) in 1978, AI is now assisting experts in selecting embryos more accurately and determining their implantation potential, according to Nathalie Massin, head of the reproduction assistance unit at the American Hospital in Paris.
The Role of Continuous Embryo Observation Systems
The American Hospital in Paris, which claims to perform over 2,300 IVF procedures annually, is equipped with an embryoscope (or time-lapse system) that continuously films the development of embryos without removing them from incubators. Anne-Claire Leprêtre, head of the reproduction assistance area at France’s public biomedicine agency, states that these continuous observation systems provide data that were not readily observable before.
Previously, the data gathered from these embryo recordings (morphology, symmetry, cell division rhythm) were used in a limited capacity. However, with AI modules, experts can now identify embryos with higher implantation potential or freezing chances, thereby reducing the number of failed attempts that could lead to spontaneous abortions.
Human Decision-Making Remains Central
“Humans will continue to make the decisions, but with this additional tool,” emphasizes Frida Entezami, co-head of the reproduction assistance center at the American Hospital in París. This hospital employs an AI system from the emerging Israeli company AIVF, currently undergoing internal validation, with the goal of reducing pregnancy cycle numbers by half.
Entezami explains that AIVF’s AI will provide a 70% probability of recommending an embryo free from genetic anomalies, a significant improvement considering that half of pre-implantation embryos are genetically abnormal currently. Nevertheless, these advancements raise questions such as handling well-qualified embryos for implantation that the algorithm flags with chromosomal anomalies.
AI’s Role in Optimizing Hormonal Treatments
AI can also assist in fine-tuning the timing and dosage of hormonal injections to optimize ovarian stimulation before egg extraction or increase the chances of locating a sperm in scarce samples. Moreover, AI algorithms are tested to verify if current observation criteria remain relevant and if additional data can refine analysis, according to Leprêtre.
“Personalized Responses”
The analysis of this extensive information and numerous cycles from various women treated with parameters of their partners or donors “could provide personalized responses” and alleviate the emotional rollercoaster often associated with these lengthy, complex, and sometimes psychologically challenging processes, as per Leprêtre.
“There’s a lot of talk about AI, but currently, it cannot do everything,” clarifies Michael Grynberg, a French obstetrician-gynecologist specializing in IVF. “We need more relevant markers because morphological criteria for eggs or sperm are insufficient.”
Key Questions and Answers
- What is the main goal of AI in assisted reproduction? The primary aim is to reduce the number of pregnancy cycles needed by half and improve embryo selection, hormonal treatment responses, and access times.
- How does AI assist in embryo selection? AI modules help identify embryos with higher implantation potential or freezing chances, reducing failed attempts.
- What role does continuous embryo observation play? Embryoscopes and time-lapse systems provide data on embryo development, which was previously unavailable or limited.
- How does AI optimize hormonal treatments? AI assists in fine-tuning the timing and dosage of hormonal injections for better ovarian stimulation and sperm detection in scarce samples.
- What are the potential challenges with AI in assisted reproduction? Questions remain about handling embryos flagged with chromosomal anomalies by AI algorithms and the need for more relevant markers beyond morphological criteria.
