A virtual speller system using SSVEP and electrooculogram

In this study a novel hybrid speller/keyboard system that combines electro-oculogram (EOG) with steady state visual evoked potential (SSVEP) is designed. Conventional EOG based speller system needs continuous eye movements for selecting a single target and it has a limitation on total number of targets. In this proposed speller, 36 targets are divided into nine groups, which includes alphabets, numbers and special characters. Target selection consists of two stages. Various eye movements (gaze, blinks, winks) are used for selecting the target groups and SSVEP is used to identify the target from the selected group. The cue guided online and free spelling tasks were performed on 10 subjects to validate the proposed system. The average classification accuracy of the proposed system is 94.16% with the information transfer rate (ITR) of 70.99 (±9.95) bits/min. For validating the proposed speller system, the classification accuracy and ITR were compared with conventional speller systems.