Analysis of speech imagery using brain connectivity estimators

Abstract

The estimation of brain connectivity allows description of the functional links established between different cortical areas during different forms of mental imagery. Speech imagery is a form of mental imagery, which refers to the activity of talking to oneself in silence. In this paper, coherence, an EEG synchronicity parameter is calculated to quantitatively analyze the concurrence of the different regions of the brain while performing speech imagery. Brain connectivity measures of speech imagery based on EEG were also investigated to understand brain function. In particular, Granger causality parameters such as Partial Directed Coherence (PDC) and Directed Transfer Function (DTF) measurements based on MVAR models are applied to multi-channel EEG data to find direction and strength of the connectivity patterns of the given speech imagery task. From the results obtained, it can be observed that there is a bilateral brain interaction of frontal and temporal brain regions andthe cross electrode coherence of the left frontal lobe was found to be high during speech production and that of the left temporal lobe was found to be high during speech imagery due to the proximity of the electrodes to the Broca's and Wernicke's area respectively. It can also be concluded that the direction of information flow from left hemisphere of the brain is more than right hemisphere of the brain using brain connectivity parameters based on MVAR models. Thus, the perceptibility of verbalizations in the brain, or in other words, speech imagery can be captured through EEG and the observations suggest that the proposed methodology is a promising non-invasive approach to study directional connectivity in the brain between mutually interconnected neural populations.