Artifacts classification in eeg signals based on temporal average statistics

EEG data contamination due to artifacts, such as eye blink, muscle activity, body movement and others pose as an issue in EEG analysis. This study aims to classify three different types of artifacts in EEG signal, namelyocular, facial muscle and hand movement using statistical features coupled with neural networks as classifier. Temporalaverages of five features are used as the feature vector for MLP classification. The experimental results for ocular, facial muscle and hand movement artifacts identification are ranging between 80% and 92%. The classification accuracy for the combination of these EEG artifacts and normal EEG of the subject for resting and eyesclose state are 86% and 96% respectivel