Premium
Single stimulus location for two inputs: A combined brain–computer interface based on Steady‐State Visual Evoked Potential (SSVEP)
Author(s) -
Wang Lu,
Zhang Zhenhao,
Han Dan,
Zhang Zhijun,
Liu Zhifang,
Liu Wei
Publication year - 2021
Publication title -
european journal of neuroscience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.346
H-Index - 206
eISSN - 1460-9568
pISSN - 0953-816X
DOI - 10.1111/ejn.15030
Subject(s) - brain–computer interface , computer science , stimulus (psychology) , visual evoked potentials , visualization , information transfer , eye movement , interface (matter) , electroencephalography , speech recognition , evoked potential , computer vision , artificial intelligence , neuroscience , psychology , cognitive psychology , telecommunications , bubble , maximum bubble pressure method , parallel computing
Brain–computer interfaces (BCI) help severely paralyzed people communicate with the outside world. One type of BCI depends on eye movements and has high information transfer (ITR) but is tiring for users and not applicable to people with eye dyskinesia. Conversely, independent BCIs enable attention shifts across visual stimuli without eye movement, but at the cost of a lower ITR. Steady‐state visual evoked potential (SSVEP) is an oscillatory brain response and typically used as BCI signal sources because of high signal‐to‐noise ratio (SNR). Considering the effect of attentional modulation on the SSVEP, we proposed the novel concept of one‐to‐two BCI to optimize existing problems, wherein the target and other stimuli shared the same location. Specifically, two spatially overlapping stimuli were displayed in the center‐of‐view field, as in the independent BCI, and participants were required to divide their attention between the right and left visual fields, as in the dependent BCI. Using three different design schemes in two experiments, we aimed to provide a new framework for BCI design by exploring the feasibility of a combined BCI that can realize a single stimulus location for two inputs. The results strongly demonstrated that, even when the targets and distractors overlapped spatially, the former evoked stronger SSVEP responses. Notably, the BCI scheme based on the object‐based attention could achieve a recognition rate as high as 83.2% and an ITR of 12.5 bits per minute. The feasibility of a one‐to‐two BCI design, which simplified the keyboard layout, reduced the attention shift, and relieved user fatigue, was established.