Neurofeedback and AI for Analyzing Child Temperament and Attention Levels

Abstract

One of the common problems among preschool children is attention ability development. It is important to detect and identify earlier the attention problems which may minimize the harmful impact of childhood disorders. The purpose of this research is to predict and analyze the attention levels of children aged 4–7. Using parental report or subjective report to analyze the children’s psychological dimensions of temperament is a common approach for temperament research, but it may be bias. Electroencephalography (EEG) is a method to illustrate the brain electrical activity. We proposed a Neurofeedback Technology (NFT) system to amalgamate the collection of EEG signals data and Behavior Style Questionnaire (BSQ) for child temperament data by applying k-means algorithm, an Artificial Intelligence (AI) unsupervised machine learning, clustering analysis method, to observe children’s attention levels. The experimental results not only infer that the value of temperament with EEG classification could be consistent, but also provide a valid way to classify attention levels in specific time period. The combination of the parental subjective report with EEG data demonstrates a novel and valuable approach for resolving child attention problems. The results facilitate earlier identification of attention problems and support better parent-child understanding and interactions.