Effects of a Dynamic Reference Frame in Mental Task Classification for EEG-Based Brain-Machine Interface

We study the effects of a moving reference frame on the classification of different mental tasks in a brain-machine interface (BMI). We use the band powers and power differences of the electroencephalogram (EEG) signals from 8 surface electrodes during 5 pre-determined mental tasks as the features for the neural network (NN) mental task classifier. We compare the NN classifier performance when the BMI user is visually subjected to a non-moving (static) reference frame (SRF) and to a moving (dynamic) reference frame (DRF) while performing the mental tasks. Results consistently show the degradation of classification accuracy for mental tasks when the user is in a DRF. Furthermore, we observe that mental tasks that require more focus are more sensitive to the SRF-DRF switch than those tasks which need less concentration when performed