Functional connectivity evaluation for infant eeg signals based on artificial neural network

Abstract

. The employment of the brain signals electroencephalography (EEG) could supply a deep intuitive understanding for infants behaviour and their alertness level within the living environment. The study of human brain through a computer-based approach has increased significantly as it aims at the understanding of infants’ mind and measures their attention towards the surrounding activities. The artificial neural network achieved a significant level of success in different fields, such as pattern classification, decision making, prediction, and adaptive control by learning from a set of data and construct weight matrices to represent the learning patterns. This research study proposes an artificial neural network based approach to predict the rightward asymmetry or leftward asymmetry which reflects higher frontal functional connectivity in the frontal right and frontal left, respectively within infant’s brain. In the traditional methods, the value of asymmetry of the frontal (FA) functional connectivity is used to determine the rightward or the leftward asymmetry while the proposed approach is trying to predict that without going through all the levels of the calculation complexity. The achieved work will supply a deep understanding into the deployment of the functional connectivity to provide information on the interactions between different brain regions.