A Brain Computer Interface Based Visual Keyboard System Using SSVEP and Electrooculogram

• Published in: Machine Learning and Metaheuristics Algorithms, and Applications Vol. 1203; pp. 64 - 74
• Main Authors: Saravanakumar, D., Ramasubba Reddy, Machireddy
• Format: Book Chapter
• Language: English
• Published: Singapore Springer 2020; Springer Singapore
• Series: Communications in Computer and Information Science
• Subjects: Brain computer interface (BCI); Electro-oculogram (EOG); Extended multivariate synchronization index (EMSI); Steady-state visual evoked potential (SSVEP); Visual keyboard
• ISBN: 9811543003; 9789811543005
• ISSN: 1865-0929; 1865-0937
• DOI: 10.1007/978-981-15-4301-2_6

This study aims to design a steady-state visual evoked potential (SSVEP) based, on-screen keyboard/speller system along with the integration of electrooculogram (EOG). The characters/targets were designed using the pattern reversal square checkerboard flickering visual stimuli. In this study, twenty-three characters were randomly selected and their corresponding visual stimuli were designed using five frequencies (6, 6.667, 7.5, 8.57 and 10 Hz). The keyboard layout was divided into nine regions and each region was identified by using the subject’s eye gaze information with the help of EOG data. The information from the EOG was used to locate the area on the visual keyboard/display, where the subject is looking. The region identification helps to use the same frequency valued visual stimuli more than once on the keyboard layout. In this proposed study, more targets were designed using less number of visual stimulus frequencies by integrating EOG with the SSVEP keyboard system. The multi-threshold algorithm and extended multivariate synchronization index (EMSI) method were used for eye gaze detection and SSVEP frequency recognition respectively. Ten healthy subjects were recruited for validating the proposed visual keyboard system.