Classification improvement and analysis of P300 responses with various inter-stimulus intervals in application to spatial visual brain-computer interface

We report on a P300 based spatial visual brain-computer interface (BCI) application improvement based on an inter-stimulus-interval (ISI) optimization. The proposed system allows for nine commands´ application using a non-invasive electroencephalography (EEG) brainwave monitoring. This paper presents the experiments results obtained by relying entirely on the visual oddball paradigm-based interaction. The visual stimuli are generated utilizing images on a computer screen arranged in a 3 × 3 matrix. The visual stimuli are used to elicit event related potentials (ERPs) with P300 components elicited to the intentional targets. The resulting ERPs are processed to extract the P300 responses in EEG features for a subsequent classification accuracy analysis. We propose to utilize a linear support vector machine (linSVM) classifier in offline EEG data post-processing analysis scenario. We discuss results of experiments conducted with five healthy users. We compare BCI accuracy results from two experimental setups with different ISI settings.