Abstract:AIM Brain-computer interface BCI based on the signals of electroencephalogram has been rapidly developed in recent years. BCI provides paralyzed individuals a new alternative method to communicate with external surroundings. A high accuracy BCI is designed using electroencephalogram EEG signals where the subjects have to think of only a single mental task. METHODS The structure of a BCI could singly be divided into two parts the front part was character extraction and classification of EEG signal the back part was combination and application with necessary hardware.BCI with high classification was designed in the front part. First EEG data characters of different metal tasks were extracted with wavelet-entropy then back propagation BP neural network was used as classifier. The obtained signal characters were used to classify different mental EEG data. RESULTS The EEG data used in this study were collected by Colorado State University Computer Science Department. There were five mental task EEG data of five subjects. Wavelet-entropy was used as signal features in EEG signal classification of two different mental tasks. The mental tasks were detected by a BP neutral network classifier. Combinations of resting baseline state and another one mental task in 4 statuses were studied for each subject. The results showed that the average classification accuracy was over 90% provided that the most suitable mental task was used. CONCLUSION Feature extraction method with wavelet-entropy can get high classification accuracy this method could be used to process mental task EEG data and wavelet-entropy would be a promising method to extract feature for BCI designs. Meanwhile this BCI design method with high accuracy can be used for classification of several mental tasks. This designed BCI which can classify several tasks would have a wide application prospect.

Classification for Different Mental Tasks Based on EEG Signals

Mental task classification for brain computer interface application

EEG Recognition Based on Multiple Types of Information by Using Wavelet Packet Transform and Neural Networks

Application of High Frequency Component in Classification of Different Mental Tasks

Mental EEG analysis based on independent component analysis

An Approach Based on BP Neural Network and Optimal F Features for Classifying Single-Trial EEG

An Extended Computer-Aided Diagnosis System for Multidomain EEG Classification

Mental EEG Analysis Based on Infomax Algorithm

An Extended Computer Aided Diagnosis System for Robust BCI Applications

Brain computer interface design of wavelet-entropy extracting mental tasks

Research of Mental EEG Classification Based on Hilbert-Huang Transform

Classifying EEG Signals Based HMM-AR

Identification and Classification of Electroencephalogram Signals Based on Independent Component Analysis

High Resolution Time-Frequency Analysis for Event-Related Electroencephalogram

CSP Based Extraction and F-Score Based Optimization of Time-Frequency Power Features for EEG Mental Task Classification

The application of Independent Component Analysis in the pattern extraction of Mental EEG

CLASSIFICATIONS OF EEG SIGNALS FOR MENTAL TASKS USING ADAPTIVE RBF NETWORK

Study on Brain-Computer Interface Based on Mental Tasks

Classification of EEG Signals Using a Multiple Kernel Learning Support Vector Machine.

Mental Task Classification Based on Adaptive Autoregressive Model and Support Vector Machine

Study on Brain Computer Interface Based on Cognitive Task