ANN Classification for the Analysis of 3D EEG Data in BBI

Abstract

In this paper, the Artificial Neural Network (ANN) algorithm is used for classifying the 3D EEG PSD in brain balanced index (BBI) is presented. The EEG signal recording was conducted on 96 healthy subjects. Development of 3D EEG models involves pre-processing of raw EEG signals and construction of spectrogram images. The maximum PSD values were extracted as features from the model. There are three indexes for balanced brain; index1, index2, index 3, index 4 and index 5. There are significant different of the EEG signals due to the BBI. Theta-θ (4-8 Hz), Delta-δ (0.5-4 Hz), Alpha-α (8–13 Hz) and Beta-β (13–30 Hz) were used as input signals for the classification model. This result has shown that the ANN classifier managed to produced accuracy rate, sensitivity rate and specificity with small classification errors for the 3D EEG PSD inputs. The overall classification accuracy of 88.89%, classification sensitivity within range 87.50% to 92.31% and classification specification within range 94.92% to 98.82% were obtained.