A Tactile Brain-Computer Interface for Virtual Wheelchair Control at Home
Matthias Eidel, Wolfgang Tröger, Martin Winterholler, T. Giesler, Andrea Kübler
Abstract
Brain-Computer Interfaces (BCI) allow for the interaction with the environment without relying on intact motor function. BCI paradigms based on visual stimulation are prevalent and highly efficient but fail once vision or gaze control are impaired. We present a single-case study of a potential end-user who operated a potentially vision independent, vibrotactile BCI. The participant in the locked-in state (LIS) trained with the BCI on a weekly basis for several months in his own home. The tactile BCI offered four movement commands which had to be selected either according to an on-screen instruction, or to navigate a virtual wheelchair through an apartment. Average accuracies of up to 67.7% were achieved. A P300 event-related potential (ERP) was consistently observed in every session, but no training effects could be demonstrated. Finally, analysis revealed a relatively high subjective workload of the paradigm.