Pre-Frontal Electrode-Based EEG Analysis for Classifying Specific Learning Disability using Support Vector Machine

Specific Learning disability (SLD) occurs when a person struggles to read, write, interpret, and spell. Reading and attending normal classes is challenging for children with SLD. Therefore, early detection and classification of SLD children from normal children is essential to improve the quality of life for SLD children. The main objective of the present study is to propose a model to classify SLD children from normal children based on identifying the dominant brain regions from their electroencephalography (EEG) findings. This study uses EEG signals from Fp1 and Fp2 regions to analyze the difference in brain activity between SLD and non-SLD children. This study included 12 SLD and 12 non-SLD children with an age group of 3–14 years. Processing of raw EEG signals, which were taken from the children, was performed using Lab Viewsoftware which involves extracting information from different frequency bands- alpha, beta, delta, gamma and theta bands. The extracted information includes the relative power and relative phase of each band. This study uses Statistical analyses such as Mann Whitney U test, Wilcoxon signed rank test and Friedman test to classify SLD children from normal children. The statistical analyses showed a significant difference between SLD and non-SLD children in Beta Phase of Fp2 region which is used in classifying the SLD children using Machine Learning approach. The classification model uses support vector machine (SVM) with linear, cubic, radial basis function (RBF) and sigmoid kernels. The Machine Learning model had a maximum accuracy of 80% using RBF-SVM classifier.