SoylentNews is people
SoylentNews
Artificial intelligence in healthcare is often a story of percentages. One 2017 study predicted AI could broadly improve patient outcomes [open, DOI: 10.7717/peerj.7702] [DX] by 30 to 40 percent. Which makes a manifold improvement in results particularly noteworthy.
In this case, according to one Israeli machine learning startup, AI has the potential to boost the success rate of in vitro fertilization (IVF) by as much as 3x compared to traditional methods. In other words, at least according to these results, couples struggling to conceive that use the right AI system could be multiple times more likely to get pregnant.
The Centers for Disease Control and Prevention defines assisted reproductive technology (ART) as the process of removing eggs from a woman's ovaries, fertilizing it with sperm and then implanting it back in the body.
The overall success rate of traditional ART is less than 30%, according to a recent study [open, DOI: 10.5455/aim.2019.27.205-211] [DX] in the journal Acta Informatica Medica.
But, says Daniella Gilboa, CEO of Tel Aviv, Israel-based AiVF—which provides an automated framework for fertility and IVF treatment—help may be on the way. (However, she also cautions against simply multiplying 3x with the 30% traditional ART success rate quoted above. "Since pregnancy is very much dependent on age and other factors, simple multiplication is not the way to compare the two methods," Gilboa says.)
Journal Reference:
Abhimanyu S. Ahuja. The impact of artificial intelligence in medicine on the future role of the physician, PeerJ (DOI: 10.7717/peerj.7702)
(Score: 2) by fakefuck39 on Friday January 01 2021, @07:51PM (2 children)
This study specifically excludes the most common method of implantation, and I believe counts the numbers wrong. They're looking at the success of an individual egg being implanted, and comparing it to pre-selecting a single egg well.
Now, I may be wrong, since not only does this study have gramaticall errors, it's not very coherently written. What I believe they're doing is this:
Multiple eggs are extracted and fertilized in a test tube. All those eggs (usually 3) are implanted, for chances of at least one taking - your body is going to select the best candidate, and very infrequently this ends up with twins. So usually, only one egg makes it into a fetus. Using the math in this study, this is a 66% failure rate. What this their method suggests is using AI to select the single egg with best chance of survival, and only implanting that egg, instead of letting the uterus make that choice.
Their method results in a higher success rate per egg. We don't give a shit about that - we care about the best success rate per procedure, and the current method of implanting multiple embryos gives a better result.
But hey, they got AI and a flowchart, and they use "predictive modeling."
(Score: 0) by Anonymous Coward on Saturday January 02 2021, @07:17PM (1 child)
We should care about the success on a per egg basis. Multiple eggs are used because if you use a number of eggs, you increase the likelihood of having one egg successfully implant. The result is that you have the occasional issues with IVF leading to multiple fetuses and ultimately babies. Having increased likelihood of an individual egg, would mean fewer eggs being tried, which would lead to increased likelihood of pregnancy and reduce the number of eggs used in the process. This would also lead to fewer eggs being used per round, so potentially more attempts for the eggs harvested.
(Score: 2) by fakefuck39 on Saturday January 02 2021, @09:45PM
I did not disagree with what you said. What I said is that the math they're using to sell this to the public is bs. Which it is, since they are presenting this as chances of success for getting pregnant, which it is not. It is however, everything you mentioned, which is an improvement, but not nearly as important as their marketing claim.