Sami Arica | PubMed Health

PubMed2006

Extraction subject-specific motor imagery time-frequency patterns for single trial EEG classification.

We introduce a new adaptive time-frequency plane feature extraction strategy for the segmentation and classification of electroencephalogram (EEG) corresponding to left and right hand motor imagery of a brain-computer interface task. The proposed algorithm adaptively segments the time axis by dividing the EEG data into non-uniform time segments over a dyadic tree. This is followed by grouping the expansion coefficients in the frequency axis in each segment. The most discriminative features are selected from the segmented time-frequency plane and fed to a linear discriminant for classification. The proposed algorithm achieved an average classification accuracy of 84.3% on six subjects by selecting the most discriminant subspaces for each one. For comparison, classification results based on an autoregressive model are also presented where the mean accuracy of the same subjects turned out to be 79.5%. Interestingly the subjects and two hemispheres of each subject are represented by distinct segmentations and features. This indicates that the proposed method can handle inter-subject variability when constructing brain-computer interfaces.

Algorithms↗

PubMed2006

Classification of single trial motor imagery EEG recordings with subject adapted non-dyadic arbitrary time-frequency tilings.

We describe a new technique for the classification of motor imagery electroencephalogram (EEG) recordings in a brain computer interface (BCI) task. The technique is based on an adaptive time-frequency analysis of EEG signals computed using local discriminant bases (LDB) derived from local cosine packets (LCP). In an offline step, the EEG data obtained from the C(3)/C(4) electrode locations of the standard 10/20 system is adaptively segmented in time, over a non-dyadic grid by maximizing the probabilistic distances between expansion coefficients corresponding to left and right hand movement imagery. This is followed by a frequency domain clustering procedure in each adapted time segment to maximize the discrimination power of the resulting time-frequency features. Then, the most discriminant features from the resulting arbitrarily segmented time-frequency plane are sorted. A principal component analysis (PCA) step is applied to reduce the dimensionality of the feature space. This reduced feature set is finally fed to a linear discriminant for classification. The online step simply computes the reduced dimensionality features determined by the offline step and feeds them to the linear discriminant. We provide experimental data to show that the method can adapt to physio-anatomical differences, subject-specific and hemisphere-specific motor imagery patterns. The algorithm was applied to all nine subjects of the BCI Competition 2002. The classification performance of the proposed algorithm varied between 70% and 92.6% across subjects using just two electrodes. The average classification accuracy was 80.6%. For comparison, we also implemented an adaptive autoregressive model based classification procedure that achieved an average error rate of 76.3% on the same subjects, and higher error rates than the proposed approach on each individual subject.

Algorithms↗