Researchers have developed artificial intelligence capable of predicting the risk of intellectual disability in premature babies.
- Researchers have developed artificial intelligence capable of predicting, from the beginning of their life, the risk of intellectual disability in premature babies.
- Ultimately, this tool could better guide doctors in their care and help parents better understand the future.
- 55,000 children are born prematurely in France each year, according to Inserm.
55,000 children are born prematurely in France each year, according to the National Institute of Health and Medical Research (Inserm), including 15% of adults (from 6 months to 7 months of pregnancy) and 5% of very premature babies (below six months of pregnancy).
Artificial intelligence predicts the risk of premature babies
Premature children can present difficulties, particularly neurological ones. These are common and can manifest as motor disorders (a delay in walking or difficulty walking), cognitive disorders (difficulties with oral or written language), attention and behavior disorders (hyperactivity, difficulties in social interactions, etc.) or even sensory, visual or hearing disorders. These babies must therefore be closely monitored medically.
Researchers from theUMC Utrecht have developed a tool that could improve this monitoring. This is an artificial intelligence (AI) capable of predicting, from the beginning of their life, the risk of intellectual disability in premature babies. Their study was published in the journal The Lancet Digital Health.
To test the reliability of this AI, scientists had it analyze the electroencephalograms (EEG) of 369 infants, taken during their first three days of life. Thus, the tool was able to distinguish, with an accuracy of around 80%, those who had a low intelligence quotient (IQ) from those who had good results at school.
Prematurity: an AI capable of answering parents’ questions
In a communicated, Maria Luisa Tataranoone of the authors of this study, judges this very important discovery, because “this means additional help, resources and care can be directed where it is most needed, from the earliest days.“
Ultimately, if this artificial intelligence validates future clinical trials, it could be useful for doctors, but also for parents. “The question they ask us most often is “What should we expect?”develops Maria Luisa Tataranno. An accurate forecast could [les soulager] or help them prepare for what awaits them.”
In the coming years, the UMC Utrecht teams will continue to develop AI. “EEG is just the beginningnotes Xiaowan Wang, other author of this study. We plan to incorporate [diverses] medical data, such as brain MRI scans, as well as data on blood oxygenation and perhaps even babies’ movements, to further refine our predictions.”
