IVF treatment has offered a wonderful opportunity to many couples who suffer from difficulties in conceiving a child. However, while the fertility treatment has meant many who may not have been able to conceive without it are able to give birth to healthy children, IVF does carry heightened risks around the likelihood of miscarriage or a still birth. Both are a devastating experience for anyone to have to live through.
In the case of heterosexual couples, IVF involves a number of the woman’s eggs being fertilised with her partner’s sperm in a laboratory environment. Very early-stage embryos are then transferred into the womb and in a successful procedure at least one attaches to the womb lining and survives. Instances of dizygotic or fraternal (non-identical) twins, or even triplets, are a relatively common occurrence in IVF-aided pregnancies because of the chance that more than one of the transferred embryos ‘takes’, after being placed into the womb.
The heightened risk arrives from the fact that, until now, doctors have lacked an effective way to assess the relative quality of the embryos chosen to be transferred into a womb. They are examined under a microscope and using time lapse imaging but this is not an effective screening method for abnormal numbers of chromosomes, something which is responsible for as many as 50% of all miscarriages.
However, the latest technology in the world of AI is now giving IVF doctors a much higher chance of helping pregnant mothers avoid the emotional distress of a miscarriage. AI algorithms tested for embryo scanning have been found to predict with 85% accuracy which embryos will survive to a live birth before they are transferred to the womb. Clinics won’t even have to own the computers or algorithms used for the embryo quality scans and will be able to upload images to the cloud with the results returned.
The Deep Learning branch of AI, which involves algorithms ‘learning’ through patterns detected based on huge volumes of cloud-stored big data, is helping improve the accuracy of many kinds of medical analysis, not only IVF. Experts believe that the potential AI analysis is showing, with the cloud-based big data and high-speed GPU processors that make Deep Learning possible only recently a reality, will lead to a new generation of medicine with the potential for huge advances to be made over coming years.