Paper Info
Title
Performance improvement of ERP-based brain–computer interface via varied geometric patterns
Abstract
Recently, many studies have been focusing on optimizing the stimulus of an event-related potential (ERP)-based brain–computer interface (BCI). However, little is known about the effectiveness when increasing the stimulus unpredictability. We investigated a new stimulus type of varied geometric pattern where both complexity and unpredictability of the stimulus are increased. The proposed and classical paradigms were compared in within-subject experiments with 16 healthy participants. Results showed that the BCI performance was significantly improved for the proposed paradigm, with an average online written symbol rate increasing by 138% comparing with that of the classical paradigm. Amplitudes of primary ERP components, such as N1, P2a, P2b, N2, were also found to be significantly enhanced with the proposed paradigm. In this paper, a novel ERP BCI paradigm with a new stimulus type of varied geometric pattern is proposed. By jointly increasing the complexity and unpredictability of the stimulus, the performance of an ERP BCI could be considerably improved.
Year
DOI
Venue
2017
10.1007/s11517-017-1671-5
Medical & Biological Engineering & Computing
Keywords
Field
DocType
Brain–computer interface,Event-related potential,P300,Paradigm
Symbol rate,Computer science,Brain–computer interface,Event-related potential,Artificial intelligence,Stimulus (physiology),Performance improvement
Journal
Volume
Issue
ISSN
55
12
1741-0444
Citations 
PageRank 
References 
0
0.34
4
Authors
2
Name
Order
Citations
PageRank
Zheng Ma101.01
Tianshuang Qiu231343.84