Resumen de Group differences in time-frequency relevant patterns for user-independent BCI applications
F. Velásquez, A. F. Zapata, Jorge I. Padilla Buriticá, José Manuel Ferrández Vicente, Germán Castellanos Domínguez
We present a comparison of two known methodologies forgroup analysis in EEG signals, which are the analysis by Group ICAon synchronization and desynchronization ERS/ERD, and brain connectivity analysis by measuring wPLI, both analyzes based on the brain synchronization information. For comparison, we have taken into account different frequency bands related to sensorimotor stimuli and time segmentation in order to overcome the nonstationarity of the EEG signal. Inaddition, we have used a threshold algorithm to reduce the dimension of the connectivity matrix, conserving the connections that are most important for both methodologies. The results obtained from the BCI competition IV-2a database show that the variable can be measured between two different measurement spaces, using the Euclidean distance, conservingspatial zones with more meaningful physiological interpretation
