2023-06-01 17:00:00
Premature birth is a major cause of infant mortality worldwide and is associated with an increased risk of disabilities in children born before the 37th week of gestation. To improve the management of these high-risk pregnancies, a team of researchers from the School of Engineering McKelvey from Washington University in St. Louis (USA) has developed an artificial intelligence (IA) able to predict premature births.
Prematurity : an AI can predict risks as early as 31 weeks
AI that relies on deep learning (i.e. the machine is able to learn by itself), analysis of clinical information such as the mother’s age, weight, presence of bleeding as well as measurements ofelectrohysterograms (EHG) 30 minutes. This involves recording the electrical activity of the uterus, a common procedure in monitoring high-risk pregnancies.
By training with data from 159 women at least 26 weeks pregnant, the tool was able to predict the risk of premature birth from the 31st week of gestation. “The deep learning algorithm performed better than other methods and provided a good way to combine data EHG with clinical information“explained Ur Goldsteinone of the researchers in a press release taken up by Neuroscience News.
A discovery that may reduce the risk of premature delivery
The American team also noticed that its algorithm was also effective with recordings EHG less than 30 minutes. The tool might thus be easier to use and possibly usable at home, according to her.
The scientists, also specify in their article published in the journal Plos Onethat they intend to continue their research to develop a device capable of recording the measurements ofEHG and to collect clinical data from a larger number of pregnant women in order to improve their prediction tool. They hope that their work will make it possible to develop applications capable of identifying premature births from the start of pregnancy in order to improve their care and reduce the risks for mothers and babies.
1685641188
#predict #premature #births
