Research on Parameter Optimization in Speech Rehabilitation System based on Brain Computer Interface

Brain-computer interface (BCI) system provides a new way of rehabilitation for patients with aphasia. 9 healthy subjects were selected to take part in this study. And EEG signals were acquired synchronously while the subjects reading Chinese characters silently. To distinguish four characters better, the EEG signal feature were selected and optimized from time-frequency domain and spatial domain. The significant time and frequency range for signal feature were selected by event related spectral perturbation (ERSP) at first. Common spatial pattern (CSP) was used to represent the spatial distribution. Different channel groups are classified by Fisher classifier to obtain the optimal channel group. The results show that spectral energy was dynamic changed in alpha and beta bands while all characters were read silently. Averaged matching accuracy of Chinese characters were improved by used the modified time and frequency range. And the matching accuracy was increased by 3.37% than used the feature from unified time and frequency range. This study contributes to the development of speech brain-computer interface, which provide new ideas for speech rehabilitation.