Individualization of data-segment-related parameters for improvement of EEG signal classification in brain-computer interface

Abstract

In electroencephalogram (EEG) modeling techniques, data segment selection is the first and still an important step. The influence of a set of data-segment-related parameters on feature extraction and classification in an EEG-based brain-computer interface (BCI) was studied. An auto search algorithm was developed to study four data-segment-related parameters in each trial of 12 subjects’ EEG. The length of data segment (LDS), the start position of data (SPD) segment, AR order, and number of trials (NT) were used to build the model. The study showed that, compared with the classification ratio (CR) without parameter selection, the CR was increased by 20% to 30% with proper selection of these data-segment-related parameters, and the optimum parameter values were subject-dependent. This suggests that the data-segment-related parameters should be individualized when building models for BCI.

Keywords

data segment / parameter selection / EEG classification / brain-computer interface (BCI)

References

[1] Wolpaw J. R., Birbaumer N., Heetderks W. J., et al. Braincomputer interface technology: A review of the first international meeting[J]. IEEE Trans Rehab Eng, 2000, 8(2): 164-173.

[2] Pfurtscheller G., Neuper C., Guger C., et al. Current trends in Graz brain-computer interface (BCI) research[J]. IEEE Trans Rehab Eng, 2000, 8(2): 216-219.

[3] Burke D. P., Kelly S. P., Chazal P. D., et al. A parametric feature extraction and classification strategy for braincomputer interfacing[J]. IEEE Trans Neural Systems and Rehab Eng, 2005, 13(1): 12-17.

[4] Schroder M, Bogdan M, Rosenstiel W. Automated EEG feature selection for brain computer interfaces[C]. In: Proceedings of the 1st International IEEE EMBS Conference on Neural Engineering. Capri Island, Italy, 2003. 20–22.

[5] Jiruska P., Proks J., Drbal O., et al. Comparison of different methods of time shift measurement in EEG[J]. Physiol Res, 2005, 54 459-465.

[6] Stastny J., Sovka P., Stancak A. EEG signal classification: Introduction to the problem[J]. Radio Engineering, 2003, 12(3): 51-55.

[7] Ince N. F., Arica S., Tewfik A. Classification of single trial motor imagery EEG recordings with subject adapted nondyadic arbitrary time-frequency tilings[J]. J Neural Eng, 2006, 3(3): 235-244.

[8] Palaniappan R. Towards optimal model order selection for autoregressive spectral analysis of mental tasks using genetic algorithm IJCSNS[J]. International Journal of Computer Science and Network Security, 2006, 6(1A): 153 162

[9] Besio W., Aakula R., Koka K., et al. Development of a tripolar concentric ring electrode for acquiring accurate Laplacian body surface potentials[J]. Annals of BME, 2006, 34(3): 426-435.

[10] Besio W., Koka K., Aakula R., et al. Tri-polar concentric ring electrode development for Laplacian electroencephalography[J]. IEEE Trans BME, 2006, 53(5): 926-933.

[11] Besio W., Cao H., Zhou P. Application of tripolar concentric electrodes and pre-feature selection algorithm for brain-computer interface[J]. IEEE Trans Neural Systems and Rehab Eng, 2008, 16(2): 191-194.

[12] Pardey J., Roberts S., Tarassenko L. A review of parametric modeling techniques for EEG analysis[J]. Med Eng Phys, 1996, 18(1): 2-11.