Design of a Visual Stimulation Module for a P300-based Brain Computer Interface that uses Pictures of Hand Movements

摘要

Brain Computer Interface (BCI) combined with functional electrical stimulation (FES) have shown great potential to enhance motor rehabilitation interventions. P300-based BCIs are generally used for generating commands for various communication and control applications. However, there are few reports of P300-based BCIs targeted for motor assistance or rehabilitation. They focus on classification performance, but do not assess the effects of the visual stimulation interface in EEG signals and evoked potentials. Here, we present the design of a visual stimulation module (VSM) for a P300-based BCI, including pictures of hand movements, arranged in matrix pattern based on the classical speller paradigm. EEG signals were recorded from a healthy subject while attending the movement pictures. Evoked potentials were found in the 350-600 ms window, and in occipital and frontal scalp locations. This was confirmed by topographic r^{2 maps, and statistical tests comparing amplitude of target and non-target epochs. Although, the VMS requires adjustments and testing with more subjects the results are encouraging. Future work includes a pilot study to assess the effect of pictures of hand movement and the variation in visual stimulation parameters in the VSM.}