DocumentCode :
663042
Title :
Unsupervîsed training of brain-computer interface systems using exnectation maximization
Author :
Speier, W. ; Knall, Jennifer ; Pouratian, Nader
Author_Institution :
Bioeng. Dept. & the Med. Imaging Inf. Group, Univ. of California, Los Angeles, Los Angeles, CA, USA
fYear :
2013
fDate :
6-8 Nov. 2013
Firstpage :
707
Lastpage :
710
Abstract :
The P300 speller is a brain-computer interface (BCI) system designed to communicate language by presenting language stimuli and detecting event related potentials in a subject´s electroencephalogram (EEG) signal. The target patient population is prone to fatigue, so reducing or removing this training step could increase the amount of time available to the subject for actual BCI use. We present an expectation maximization approach that trains the classifier in an unsupervised manner. A general classifier is created from a set of multiple subjects and it is then refined using the subject´s unlabeled data and knowledge from the language domain. The method was tested offline on a data set of 15 healthy subjects and achieved similar performance to fully supervised methods for all subjects. This suggests that this method could be used in the place of the training step for BCI systems.
Keywords :
bioelectric potentials; brain-computer interfaces; electroencephalography; expectation-maximisation algorithm; hidden Markov models; medical computing; unsupervised learning; BCI system; EEG signal; P300 speller; brain-computer interface system; electroencephalogram signal; event related potentials; expectation maximization approach; fatigue; full supervised methods; language domain; language stimuli; target patient population; unsupervised training; Accuracy; Algorithm design and analysis; Brain-computer interfaces; Classification algorithms; Hidden Markov models; Training; Viterbi algorithm;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Engineering (NER), 2013 6th International IEEE/EMBS Conference on
Conference_Location :
San Diego, CA
ISSN :
1948-3546
Type :
conf
DOI :
10.1109/NER.2013.6696032
Filename :
6696032
Link To Document :
بازگشت