Title: Feature extraction and analysis of overt and covert EEG signals with speller devices
Authors: Mridu Sahu; Saumya Vishwal; S. Usha Srivalli; Sneha Shukla; Sanjivani Shantaiya
Addresses: Department of Information Technology, National Institute of Technology, Raipur, Chhattisgarh, India ' Department of Information Technology, National Institute of Technology, Raipur, Chhattisgarh, India ' Department of Information Technology, National Institute of Technology, Raipur, Chhattisgarh, India ' Department of Information Technology, National Institute of Technology, Raipur, Chhattisgarh, India ' Department of Information Technology, National Institute of Technology, Raipur, Chhattisgarh, India
Abstract: Brain computer interfaces (BCIs) are used by motor neuron disease patients for communication. The brain electroencephalogram signals, detected and recorded by BCI devices are used for analysis. The P300 component of event related potentials ERP used for the detection of attention towards a character is a popular approach. The first most successful speller device was proposed by Farewell and Donchin, named P300 speller. The device consists of 6 × 6 matrix of alpha-numeric with rows and columns randomly intensified, in order to generate a stimulus in the users' brain for efficient use of BCI. In the presented work, overt and covert attention are recognised. Overt attention is being looked at with eye movement and covert attention is not associated with eye movement. Geometric speller has been used in this study; a comparative study of GeoSpeller and P300 speller is showed in covert and overt attention for analysis of both the methods and their usability.
Keywords: geometric speller; P300 speller; overt and covert attention; brain computer interfaces; BCIs; electroencephalography; motor neuron disease; event related potentials; ERP; overt; covert.
DOI: 10.1504/IJAIP.2023.135026
International Journal of Advanced Intelligence Paradigms, 2023 Vol.26 No.2, pp.124 - 142
Received: 15 Sep 2017
Accepted: 10 Mar 2018
Published online: 28 Nov 2023 *