Growth retardation in premature babies can be predicted very early –

It is known that the cognitive functions of the brain do not develop as well in children born prematurely as in children of the same age who were born at full term. But it was difficult to predict at the individual level which children were most at risk for developmental delay and would therefore benefit most from preventive measures.

Artificial intelligence has sorted out the risk factors

Based on information in the newborn care registry, KI researchers mapped a wide range of conditions during pregnancy, birth and newborn care. It concerns Swedish babies born in weeks 22 to 31 of pregnancy. Using machine learning, artificial intelligence, researchers have sorted out the most important risk factors for delayed cognitive development.

In total, the study included 1,062 very premature babies with an average birth weight of 880 grams. They were all psychologically tested for cognition at the age of two.

Prolonged ventilator care is a risk factor

Of all the children examined, about a fifth at age 2 showed signs of delayed cognitive development. In addition to known risk factors such as low birth weight, male gender, and brain hemorrhage in newborns, researchers can now see other factors that could be linked to uniform growth. These problems include a non-Norse home language, long-term ventilator treatment, and failure to breastfeed upon discharge from neonatal care.

– This is an important step in order to design preventive interventions for children who are actually most in need regarding discharge from neonatal intensive care, says Mikael Norman, Professor of Pediatrics at Karolinska Institutet.

Effective help is available today, but it is often resource-intensive and therefore may not be provided to everyone. Therefore, new forecasting tools such as the one we have developed are needed.

Read alsoDifficult childhood is the cause of many diseases

Scientific material:

Predicting 2-year cognitive outcomes in preterm infants using machine learning methodsjam.

communication:

Mikael Norman, Professor of Pediatrics at the Department of Clinical Sciences, Intervention and Technology at Karolinska Institutet
[email protected]