17-10-2016, 09:57 AM
1459324554-paper.docx (Size: 12.55 KB / Downloads: 5)
ABSTRACT
Brain connectivity parameters help establishing the functional connectivity between different regions of the brain during mental imagery. Imagined speech, a type of mental imagery, is the process of saying a word or phrase without any vocalization or movement of lips, tongues or jaws. In this paper, the regions of the brain associated with covert and overt speech processes of Consonant and vowel combinations are analysed using brain connectivity estimators.Coherence, a synchronicity estimator, is estimated to analyse the degree of association between different regions of the brain during imagined and normal speech. To analyse the direction flow in the brain regions during the mentioned processes, Granger causality estimators such as Partial directed coherence (PDC) and Directed transfer function (DTF) measurements, based on multivariate autoregressive model (MVAR),are applied to multi-channel EEG data acquired. Coherence results indicate that frontal and temporal regions of the left hemisphere are more active than that of the right hemisphere. In specific, Beta and Gamma frequency bands showed more activity than other bands.
Keywords- Imagined speech, brain connectivity estimators, electroencephalography: EEG coherence; Partial directed coherence; Directed transfer function.