Artificial Intelligence Based 4 Axis Direction Control with Electroencephalography Signal

Abstract

Diseases such as paralysis caused by cerebrospinal cord injury and Amyotrophic Lateral Sclerosis (ALS) restrict people’s basic movements. Although these diseases, which are caused by nervous system problems, prevent the basic movements of people, they do not prevent brain activities, which allows EEG signals to be received. Electroencephalography (EEG) signals received in the study are classified as basic directional movements to make people’s lives easier. Before the classification process, the EEG signals were filtered because of the noise they contain. In order to make the classification more accurate and detailed after the filter, first epoching and then feature extraction were performed. With the extracted features, the four-axis control process was performed with a classifier called a support vector machine (SVM). For higher accuracy during classification, data are grouped as right-left and more or less. A grouped one was chosen as training and the other as a test. Thus, it is aimed to see how much the results of the study will change. Performance evaluation criteria were obtained from the data classified using the SVM algorithm. The evaluation of the results was made on the basis of the accuracy value, which is one of the performance criteria. As a result of the evaluation, the highest accuracy rate was %57 in the classification made using right-left grouped data as education, while the highest accuracy rate was %52 in the data grouped as up-down.