Visualization of ERD/ERS on leg Motor Imagery

抄録

Brain computer Interface (BCI) is the interface that is able to control the computers or machines without physical movement by reading User’s intent from specific electroencephalogram (EEG) pattern. Event Related Desynchronization (ERD) / Event Related Synchronization (ERS) is one of the specific EEG pattern. ERD is caused the decrease of EEG frequency power in alpha (8-12Hz) and beta (12-30Hz) bands on the motor area related to the body parts by preparing movement, actual movement, or Motor Imagery (MI). In contrast to ERD, ERS is caused the increase of EEG frequency power in a similar way and bands. There are many researches of BCI using ERD/ERS because this is useful as rehabilitation and logical switch information in the controller, etc. We also have the research significance to realize the interface for patients can’t walk or VR-walking machine using BCI, therefore, leg MI was used about the kind of MI. The BCI related to leg get close to realize by enhancing the detection accuracy on leg MI. To enhance the detection accuracy, we focused on quantifying ERD/ERS. However, current quantifying is not suitable for online analysis because analysist need to set reference data by selecting the data ERD/ERS appeared. Therefore, ERD/ERS detection algorithm was proposed for online analysis in this research. On the proposing that algorithm, we used alpha band (8-12Hz).