JEDDAH: The University of Jeddah has been studying the possibility of early prediction of cerebral palsy in infants with hypoxia during birth by analyzing brain signals and utilizing machine learning algorithms.
Modern medicine currently uses neurological tests to diagnose cerebral palsy after symptoms first appear in a child.
During the research, a set of measures that reflect the complexity of signals and measure functional brain connectivity were calculated.
The results of the analysis showed generalized electroencephalogram, or EEG, abnormalities in infants with cerebral palsy compared to their healthy counterparts.
In addition, the measures were included in a set of machine learning algorithms which tested the proposed model’s ability to classify a group of 26 infants. Six of them showed symptoms of cerebral palsy at the age of two years, according to neurological tests, and the remaining 20 did not develop the disorder.
A good performance of the proposed model was obtained with a classification accuracy of 84.6 percent, sensitivity of 83 percent, specificity of 85 percent, and area under the curve of 0.87.
The study demonstrated that the brain function characteristics measured were able to successfully distinguish between the two groups of infants, potentially indicating future use in clinical applications as biomarkers for detecting cerebral palsy.