Predicting the Level of Neurological Deficit for Children with Stroke Based on Mathematical Modeling

This article discusses the development of mathematical models for predicting the severity of neurological symptoms in children with ischemic stroke. To do this, it was established which parameters from the data of children anamnesis and the results of their examinations can have a significant effect on the severity of neurological symptoms at the time of discharge of the child from the hospital and one year after the end of treatment. Based on the use of the method of group accounting of arguments and neural networks, mathematical models have been constructed to make a prediction of the values of the level of neurological symptoms. The estimation of the adequacy of the constructed models and their comparison between them has been carried out, which showed that the models based on the method of group accounting of arguments showed higher accuracy of the forecast than models based on neural networks.