CLASSIFICATION OF ELECTROENCEPHALOGRAM SIGNALS

Brain Computer Interface (BCI) employs a mechanism that creates a communication between the brain and the exterior world using the brain signals. In other words, BCI creates a way to connect the brain with the outside environment. First of all, the brain signals are recorded and then it is handled to transform brain activity to its corresponding commands. Basically, there are two kinds of BCI systems, Invasive and Noninvasive, which assist the client in controlling many types of applications. Such ability can become so helpful for patients who are suffering from severemotor function issue. The methodology that makes BCIs have a high performance based on signal processing approach is utilized by feature extraction and translation of Electroencephalogram (EEG) patterns. In this work Principle Component Analysis (PCA) is used to remove artifacts and also power spectral density (PSD) is extracted from sensorimotor rhythms which are considered as a feature. This will become a foundation for Support Vector Machine (SVM) translation algorithm of EEG patterns. This work has been analyzed and evaluated using computer by means of MATLAB R2016a. The proposed approach gave a result equal to 98.2%.