Motor imagery classification using multiresolution analysis and sparse representation of EEG signals

Classifying motor imagery brain signals where the signals are obtained based on imagined movement of the limbs is a major, yet very challenging, step in developing Brain Computer Interfaces (BCIs). Of primary importance is to use less data and computationally efficient algorithms to support real-time BCI. To this end, an algorithm that exploits the sparse characteristics of EEGs is proposed to classify these signals. Different feature vectors are extracted based on the energies in different frequency sub-bands of the Wavelet Packet decomposition of EEG trials recorded only by five electrodes near the sensorimotor cortex. Features from a small spatial region are approximated by a sparse linear combination of few atoms from a multi-class dictionary constructed from the wavelet features of the projected EEG training signals for each class. This is used to classify the signals based on the pattern of their sparse representation using a minimum-residual decision rule. The results obtained from real data demonstrate that the combination of energy and entropy features enables efficient classification of motor imagery EEG trials related to hand and foot movements. Also, the decomposition used is shown to provide a frequency resolution consistent with the existence of different levels within the alpha band.