Recognition of Visual Evoked Potential Responses Containing Cognitive Component (P300) Using Reflex Fuzzy Min-Max Neural Network

This paper proposes a system to recognize Visual Evoked Potentials (VHP) responses related to cognitive component P300. Recognition of such response is essential in Brain-Computer Interface systems. The transformed (Time-Frequency) domain features of average and single trial VHP signals were extracted for classification. Reflex Fuzzy Min-Max Neural Network is used for this purpose. Experimental results show that features of average VEP signals gives better recognition accuracy compared to single trial VEP signals. Moreover it is observed that the transformed domain post-stimulus coefficients on approximate level are sufficient for the discrimination process.