A Hybrid Brain-Computer Interface Based on the Fusion of P300 and SSVEP Scores

The present study proposes a hybrid brain-computer interface (BCI) with 64 selectable items based on the fusion of P300 and steady-state visually evoked potential (SSVEP) brain signals. With this approach, row/column (RC) P300 and two-step SSVEP paradigms were integrated to create two hybrid paradig...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on neural systems and rehabilitation engineering 2015-07, Vol.23 (4), p.693-701
Main Authors: Yin, Erwei, Zeyl, Timothy, Saab, Rami, Chau, Tom, Hu, Dewen, Zhou, Zongtan
Format: Article
Language:English
Subjects:
Accuracy
Adolescent
Adult
Ash
Bayes Theorem
Brain-computer interface (BCI)
Brain-Computer Interfaces
Discriminant Analysis
Educational institutions
electro encephalogram (EEG)
Electroencephalography
Equipment Design
Event-Related Potentials, P300 - physiology
Evoked Potentials, Somatosensory - physiology
Evoked Potentials, Visual - physiology
Feature extraction
Female
Humans
Learning Curve
Male
P300
Psychomotor Performance
score fusion
steady-state visually evoked potential (SSVEP)
Support Vector Machine
Young Adult
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The present study proposes a hybrid brain-computer interface (BCI) with 64 selectable items based on the fusion of P300 and steady-state visually evoked potential (SSVEP) brain signals. With this approach, row/column (RC) P300 and two-step SSVEP paradigms were integrated to create two hybrid paradigms, which we denote as the double RC (DRC) and 4-D spellers. In each hybrid paradigm, the target is simultaneously detected based on both P300 and SSVEP potentials as measured by the electroencephalogram. We further proposed a maximum-probability estimation (MPE) fusion approach to combine the P300 and SSVEP on a score level and compared this approach to other approaches based on linear discriminant analysis, a naïve Bayes classifier, and support vector machines. The experimental results obtained from thirteen participants indicated that the 4-D hybrid paradigm outperformed the DRC paradigm and that the MPE fusion achieved higher accuracy compared with the other approaches. Importantly, 12 of the 13 participants, using the 4-D paradigm achieved an accuracy of over 90% and the average accuracy was 95.18%. These promising results suggest that the proposed hybrid BCI system could be used in the design of a high-performance BCI-based keyboard.
ISSN:1534-4320
1558-0210
DOI:10.1109/TNSRE.2015.2403270