Analysis and visualization of ERD and ERS with adapted local cosine transform

We apply adapted local cosine transform to movement EEG, which is a local spectral representation of the signal. It includes best basis method obtained by entropy minimization. The algorithm yields to adaptive time segmentation, where these segments correspond to ERD and ERS events. The algorithm provides short segments in ERD and ERD to ERS transitions. Then we use averaged DCT coefficients to extract the ERD ERS structure of a person.