DocumentCode
2626574
Title
Discrimination of "Yes" and "No" Responses by Auditory Stimuli Multiple-choice Questions in Human EEG
Author
Seo, Ssanghee ; Chen, Haifeng ; Ye, Donghee ; Lee, Jungtae
Author_Institution
Pusan Nat. Univ., Busan
fYear
2007
fDate
21-23 Nov. 2007
Firstpage
1127
Lastpage
1133
Abstract
Patients with severe physical disorders but who are cognitively intact find it difficult to communicate with the outside world. Researches on the brain-computer interface (BCI) are aimed to help such patients. This study analyzes the characteristics of brainwaves and classifies between "yes" and "no" responses when a subject thinks of a "yes" or "no" response to an auditory stimuli question. Experiment using auditory stimuli of the form of multiple-choice questions is performed. We discriminate between the activities of the brainwaves of the "yes" and "no" responses using the alpha power. Nineteen healthy subjects (7 females and 12 males), aged between 20 and 35 years, volunteered to participate in the experiments. In the experiment, for 17 out of 19 subjects (89.4%), the average alpha power for the "yes" response is greater than that for the "no" response. Also, Fisher\´s discriminant is performed to classify what the subject responses based on the average alpha power. 88.9% of original grouped cases (38 cases) and 83.3% of ungrouped cases (6 cases) are correctly classified.
Keywords
auditory evoked potentials; electroencephalography; handicapped aids; neuromuscular stimulation; user interfaces; Fisher discriminant analysis; auditory stimuli multiple-choice questions; brain-computer interface; brainwaves; human EEG; power analysis; Aging; Brain computer interfaces; Computer science; Costs; Electroencephalography; Enterprise resource planning; Frequency; Humans; Information technology; Linear predictive coding;
fLanguage
English
Publisher
ieee
Conference_Titel
Convergence Information Technology, 2007. International Conference on
Conference_Location
Gyeongju
Print_ISBN
0-7695-3038-9
Type
conf
DOI
10.1109/ICCIT.2007.58
Filename
4420409
Link To Document