Classification of dyslexic and normal children during resting condition using KDE and MLP

Dyslexia is a specific reading disability. It can be characterized by a severe difficulty in reading, learning, spelling, memorizing as well as sequencing activities. In this work, the participants´ electroencephalogram (EEG) signals were monitored during resting situation. These signals are captured from the scalp of each subject to measure the brain activities during both eyes opened and eye closed scenarios. Features from the EEG signals were extracted using the Kernel Density Estimation (KDE) and classified using the Multilayer Perceptron (MLP). Due to the large number of features extracted, relevant features are then selected by grouping various spectral components and eliminating irrelevant features. For a comparison purpose, brain signals of three children who are diagnosed of having dyslexia by medical practitioners (denoted as dyslexic) and the other three children diagnosed otherwise (denoted as normal) are used. Experimental results shown that there is a clear distinction between dyslexic and normal children during both eyes closed and eyes opened scenario. Hence, further works can be extended for early intervention in such a way that these children can be further assisted to cope with their learning experience.