They develop AI-based technology to detect babies at risk of developmental delay
The delay in neurocognitive development affects around 20% of babies and children. These are disorders that can make learning difficult and, in some cases, lead to chronic disabilities, but their symptoms do not usually appear until preschool age. A project from the University of Barcelona (UB) has received a 150,000-euro grant from the Knowledge Industry program of the Generalitat de Catalunya to develop a device that uses artificial intelligence (AI) algorithms applied to electroencephalograms (EEG) to predict the risk of suffering this delay, thus allowing intervention to begin before the first signs of the problem appear.
“Our solution could save a significant number of children from the onset of symptoms and their subsequent school failure, as well as the suffering of families and the financial burden that special educational needs entail,” explains Dr. Carles Escera, professor of Cognitive Neuroscience at the Faculty of Psychology and researcher at the UB Institute of Neuroscience and the Sant Joan de Déu Research Institute, who leads the project.
A fast, painless and non-invasive technique
The innovative solution, which already received a Llavor grant from the same program of the Generalitat de Catalunya, is based on the recording of the baby’s electroencephalogram in response to a speech sound, a fast, painless and non-invasive technique. During the last few years, researchers have identified a brain response as a biomarker of risk for neurocognitive delay from the study of babies born small for their gestational age, who are at high risk of suffering from these problems.
This brain biomarker, recorded in the electroencephalogram, is interpreted using an AI algorithm that has been developed with a database that includes more than 600 records of babies at various stages (0, 6, 12 and 21 months) , as well as other data such as neuropsychological test results and genetic information, with the aim of improving the predictive accuracy of the technology.
In this way, the algorithm is able to indicate what will be the risk of delay in the baby’s neurocognitive development at 21 months of age. “The first years of life are crucial in the learning process of babies. Thus, predicting the risk of neurodevelopmental delay at birth and applying an early prophylactic intervention can drastically reduce the prevalence of this problem which is usually detected incidentally”, explains the researcher.
The new project, with file number 600401, will allow the implementation of the algorithm designed to predict neuropsychological results to be concluded and the integration into an initial prototype to begin preclinical trials of the technology. These grants from the Generalitat’s Knowledge Industry program are intended for obtaining prototypes and for the valorization and transfer of research results generated by research teams in Catalonia.