Spatio-spectral & temporal parameter searching using class correlation analysis and particle swarm optimization for a brain computer interface

Distinct features play a vital role in enabling a computer to associate different electroencephalogram (EEG) signals to different brain states. To ease the workload on the feature extractor and enhance separability between different brain states, numerous parameters, such as separable frequency bands, data acquisition channels and time point of maximum separability are chosen explicit to each subject. Recent research has shown that using subject specific parameters for the extraction of invariant characteristics specific to each brain state can significantly improve the performance and accuracy of a brain-computer interface (BCI). This paper focuses on developing a fast autonomous user-specific tuned BCI system using particle swarm optimization (PSO) to search for optimal parameter combination based on the analysis of the correlation between different classes i.e., the R-squared (R2) correlation coefficient rather than assessing overall systems performance via performance measure such as classification accuracy. Experimental results utilizing eight subjects are presented which demonstrate the effectiveness of the proposed methods for fast & efficient user-specific tuned BCI system.