1D-CNN architectures for EEG classification with motor imagery input of eyes open and eyes closed conditions
by P. Nagabushanam; S. Thomas George; M.S.P. Subathra; S. Radha
International Journal of Intelligence and Sustainable Computing (IJISC), Vol. 1, No. 3, 2021

Abstract: Neuro cognitive performance is an interesting area which deals with normal and abnormal conditions of a person in his thinking capabilities like driving, reaction to the events happening around him. The study of such information is called sleep scoring or sleep stages classification. If it is abnormal for a person, it further leads to insomnia, hypersomnia, epilepsy disorders. Hence, early stage detection of such abnormalities may help in treating the person in time and avoiding major problems ahead. Sleep stage classification may be 2-way, 3-way, 4-way or 5-way based on type of EEG captured from the person. In this paper, we have proposed various 1D-CNN architectures for 2-way sleep stage classification for motor imagery EEG captured during eyes open and eyes closed conditions. Implementation is done in python keras in Jupiter notebook, anaconda. Among the proposed models, 1D-CNN with four layers give better accuracy of 75.753%. In general, sleep stages classification give accuracy less than 80% and also it may be based on quality of EEG input considered.

Online publication date: Mon, 22-Nov-2021

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Intelligence and Sustainable Computing (IJISC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com