An intelligent Spelling Framework Based on Brain Computer Interface

Brain-computer Interface (BCI) aims to enhance the quality of life for all humans. Spelling is one of BCI applications that is used to type numbers, characters, words, or sentences by recording the user's brain activity. In this paper, A BCI speller framework based on converting mental activity is presented. Such framework uses Independent Component Analysis (ICA) and Auto Regressive (AR) for preprocessing and feature extraction respectively. Both of Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) are utilized at the classification phase. Several experiments have been conducted by four subjects using the pre-described framework achieved high average accuracy of 94.38% for KNN with value of K=9. The performance results have shown that converting mental activity can be used as a mean for spelling applications.