Computer Vision with AI May Revolutionize IVF Treatment

The novel technique developed by researchers at IVIRMA Valencia eliminates the requirement of invasive cell biopsy in IVF treatment by utilising computer vision and AI references of cell activity observed through time-lapse imaging. The innovative method can efficiently distinguish between euploid and aneuploid embryos, determining which ones will be optimal for IVF treatment, the Innovation News Network reported.

Examining the chromosomal content of embryos has become a routine and controversial method used in many fertility clinics to determine the ideal candidates for maximising birth rates from IVF treatment. Embryos containing a normal number of chromosomes – known as euploid – have the best chance of implanting in the uterus and instigating pregnancy, whereas abnormal embryos – aneuploid, do not have a chance. Methods that investigate embryos for aneuploidy require cell samples taken through biopsy from the embryo, which has raised safety concerns and prompted scientists worldwide to develop non-invasive methods.

Over the last ten years, embryo growth has been visualised by time-lapse technology, which produces an image for each stage of the embryo’s development until it becomes a blastocyst and is ready to be transferred to the uterus. Despite this, time-lapse imaging is not able to analyse the chromosomal status of an embryo, with computer vision with AI possibly providing the perfect solution.

Chromosomally normal embryos begin developing as blastocysts at an earlier point than aneuploid embryos, which can accurately be recognised through computer vision by the microscopic measurements of the cells’ edges, making it a precise method of calculating the number of cells and cell cycle of the blastomeres – the cells that form the embryo. The team combined this finding with computer vision-based measurements of cell edges in time-lapse videos of 111 euploid and 120 aneuploid embryos, discovering that aneuploid embryos reach their blastocyst stage earlier than the euploid embryos.

Marcos Meseguer, the study director, said: “Our results show for the first time that an AI-based system can precisely measure microscopic cell edges in the dividing embryo, which allowed us to distinguish between euploid and aneuploid embryos.

“Our early results had shown that euploid and aneuploid embryos are visually distinct, significantly enough to merit further computer vision investigation and to test if a non-invasive PGT-A test could conceivably match the results of current invasive methods – without the cost and damage to the embryo that the invasive methods might cause. We used the measurement of cell edges as a proxy for cell activity (which include DNA replication and cell division) and achieved 73% sensitivity and specificity in our results.”

The researchers have reiterated that they will need to conduct further investigations to authenticate the algorithms in larger datasets, as their model classifies mosaic embryos as abnormal, even though they are believed to be suitable for pregnancy. Furthermore, they have explained that non-invasive methods experimented with so far have not produced results as accurate as conventional biopsy methods.

Meseguer said: “Our present algorithm faces the same situation. Our prediction capability is still limited, in which case our models could only be applied to those patients who do not require genetic testing according to a pre-defined medical indication. So, our test so far could only be used to reduce the risk of selecting a chromosomally abnormal embryo for transfer.”

Nevertheless, this novel computer vision approach is both efficient and economical in comparison to other non-invasive procedures, can be built at home, and may potentially become the most efficient way of examining the aneuploidy of embryos.

“These other non-invasive results take several days to produce because of the genetic analysis, which forces patients to freeze all their embryos and delay their infertility treatment,” said Meseguer.

4155/v