A co-adaptive training paradigm for motor imagery based brain-computer interface

Authors:
Bin Xia;Qingmei Zhang;Hong Xie;Shihua Li;Jie Li;Lianghua He
Affiliations:
Information Engineering College, Shanghai Maritime University, Shanghai, China;Information Engineering College, Shanghai Maritime University, Shanghai, China;Information Engineering College, Shanghai Maritime University, Shanghai, China;Information Engineering College, Shanghai Maritime University, Shanghai, China;College of Electronics and Information Engineering, Tongji University, Shanghai, China;College of Electronics and Information Engineering, Tongji University, Shanghai, China
Venue:
ISNN'12 Proceedings of the 9th international conference on Advances in Neural Networks - Volume Part I
Year:
2012

Citing 3
Cited 0

The nature of statistical learning theory

The nature of statistical learning theory
Machine learning and applications for brain-computer interfacing

Proceedings of the 2007 conference on Human interface: Part I
Machine-learning-based coadaptive calibration for brain-computer interfaces

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

In motor imagery based Brain-Computer Interface (MI-BCI), subjects should be trained to learn how to modulate the rhythm of EEG for a long time. In previous works, more researchers focused on adaptive BCI system and a few works studied neurofeedback-based subjects training. To achieve high training performance, system self-adaption and subjects training were considered simultaneously in recent works. In this work, we present a co-adaptive training paradigm which includes subjects training and BCI system training. For subjects training, we present a neurofeedback-based training paradigm applying the strength information of motor imagery. In system training, the classifier model is run-by-run updated by selecting good features from EEG data of several previous runs. The online and offline analysis demonstrate that the proposed training paradigm can achieve a better training performance than normal training paradigm.