Addi"ional informa"ion is available a" "he end of "he chap"er h""p://dx.doi.org/10.5772/55805 . IntroductionThis chapter will propose a new paradigm for single-trial-electroencephalogram EEG -based "rain-Computer Interfaces "CIs with motor imagery MI [ ] tasks. "mong such "CIs, the sensorimotor rhythm SMR -based ones, when using common spatial patterns CSPs , require features over broad frequency bands, such as mu, beta and gamma rhythms [ ]. Therefore, very high-dimensional feature vectors and continuous-valued patterns necessary for spatiotemporally checking the features [ , ] could yield an enormous amount of data and much computational time [ ]. So, various data reduction such as downsampling [ , ] and optimal EEG channel configuration [ , , ] have been investigated for the "CIs.The present method consists of the categorization of single-trial EEGs as data reduction, and the classifiers for the categorical data. is realized by equivalent current dipole source localization ECDL after independent component analysis IC" . For , we have been applying both Hayashi's second method of quantification H MQ and "ayesian network model "NM to the ECDL-based categorical data. For the former, we have obtained the good accuracy, for example, the accuracy average across all the ten subjects for left-and right-hand imageries in each -trial validation was more than % [ ].This chapter addresses itself to the single-trial-EEG-based "CI using the "NM and to the generalization to dynamic "NM D"NM because of the time-varying functional networks in the brain. For the purposes, two experiments were conducted to obtain single-trial EEGs scalp-© 2013 Yamazaki e" al.; licensee InTech. This is an open access ar"icle dis"rib""ed "nder "he "erms of "he Crea"ive Commons A""rib""ion License (h""p://crea"ivecommons.org/licenses/by/3.0), which permi"s "nres"ric"ed "se, dis"rib""ion, and reprod"c"ion in any medi"m, provided "he original work is properly ci"ed.recorded during the MI tasks and movement-related potentials MRPs including the "ereit-Recently, neuroscience has been attempting to take in various methodologies in network science, because the brain could be considered to be a kind of complex systems forming networks of interacting components, and the collective actions of the components, that is, individual neurons, linked by a dense web of intricate connectivity [ ]. In addition to the network approach, one of the applied researches in neuroscience, the "CI, has extensively received probabilistic approaches whose aims are mainly two. One is to cope with nonstationarities in EEG signals such intertrial and intersubject variations, and the other to incorporate time-varying brain states and uncertainties into "CI design. For the former aim, adaptive classifications were executed by Kalman filtering [ , ], while the D"N achieved the latter one [ ]. Micheloyannis et al. [ ] analyzed multi-channel EEGs using graph theory. However, because the nodes are electrode positions, they have few functional meanings. In addition, all the above met...
SUMMARYWe have proposed a new Bayesian network model (BNM) framework for single-trial-EEG-based Brain-Computer Interface (BCI). The BNM was constructed in the following. In order to discriminate between left and right hands to be imaged from single-trial EEGs measured during the movement imagery tasks, the BNM has the following three steps: (1) independent component analysis (ICA) for each of the singletrial EEGs; (2) equivalent current dipole source localization (ECDL) for projections of each IC on the scalp surface; (3) BNM construction using the ECDL results. The BNMs were composed of nodes and edges which correspond to the brain sites where ECDs are located, and their connections, respectively. The connections were quantified as node activities by conditional probabilities calculated by probabilistic inference in each trial. The BNM-based BCI is compared with the common spatial pattern (CSP) method. For ten healthy subjects, there was no significant difference between the two methods. Our BNM might reflect each subject's strategy for task execution.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.