Towards an sEEG-based BCI using code-modulated VEP: A case study showing the influence of electrode location on decoding efficiency

Abstract

Today, one of the fastest non-implanted brain-computer interfaces (BCIs) for communication and control uses the code-modulated visual evoked potential (cVEP) recorded with electroencephalography (EEG) (Martínez-Cagigal et al., 2021). The c-VEP is a distinct pattern of brain activity triggered by rapid pseudo-random visual stimulation and is strongest over the visual cortex. We present the first evaluation of an implanted c-VEP BCI for communication using stereoelectroencephalography (sEEG). Circumventing the effortful capping procedure of EEG, a chronic sEEG implant could possibly provide higher usability for daily life settings than EEG. Reliability may also improve, as sEEG signals are expected to show a higher signal-to-noise ratio (SNR) and a lower sensitivity to artifacts than EEG.