Reasearch On Hybrid Brain Computer Interface Based On SSVEP And Motor Imagery

Brain-computer interface(BCI) is a kind of system enabling brain to communicate with external environment, which is independent of peripheral nerves and muscular pathways. BCI systems acquire signals from the brain, classify them with signal processing algorithms and translate them into control commands for external devices. Therefore, BCI systems have wide application prospect in the field of rehabilitation. At present, one of the main problems about blocking BCI systems into practical application is that the number of electroencephalography(EEG) classification is not enough. If the BCI systems are only based on spontaneous EEG or evoked EEG, the problem can not be dealt well. This thesis is about the research on hybrid BCI based on spontaneous EEG and evoked EEG, which is an effective way to solve the problem.This thesis describes the hybrid BCI system based on Steady-State Visual Evoked Potentials(SSVEP) and Motor Imagery(MI), which combines spontaneous EEG and evoked EEG. The features of them are extracted and the number of classification is increased, which propels the BCI toward our life more closely. The main work has been listed as below:1. The design of experiment. Three paradigms are designed. It includes the SSVEP paradigm, the MI paradigm and the hybrid paradigm combining SSVEP and MI. The SSVEP paradigm researches the factors influencing the SSVEP, such as the stimulator’s generation property, flickering frequency, color and size. The MI paradigm researches the proper way to imagine. The hybrid paradigm uses the best factors of the SSVEP paradigm and the MI paradigm to research the influence on the number of classification.2. The extraction of EEG features. Firstly, the EEG pre-processing is used in the EEG data analysis. Then, the Fast Fourier Transform Algorithm(FFT) and r2 Algorithm are used to extract features of different types of EEG.3. The analysis of pattern recognition. The least amount of electrode features are used, and Support Vector Machine(SVM) method is used to classify the EEG features, which saves the time of calculation and improves the ability of on-line analysis.4. The design of on-line system. The on-line system based on hybrid BCI is designed according to the basic daily need of patients. With the system, the user can send simple control commands with brain to realize the communication with our environment.