An EEG Brain-Computer Interface to Classify Motor Imagery Signals

Abstract

Considering the increase in life expectancy, people started to invest in technologies capable of improving the quality of life. One of these technologies is the Brain-Machine Interface. Combined with EEG signals, this technique may allow individuals with some motor disabilities to perform activities of daily living. Motor Imagery came up as an important tool to support this population. So they may send commands to external devices by using their brain voluntary activity. In this chapter, the performance of an Imagery EEG-based BCI engine was accessed by applyingWavelet transform to the signals and extracting metrics used to describe digital signals. We used signals from the motor imagery of the right hand, left hand and foot movements. Different intelligent classifiers were tested. We achieved results greater than 99% of accuracy and Kappa above 0.99. The method is promising and can be used for future evaluations with several individuals to verify reproducibility.