Decoding motor imagery hand direction in brain computer interface from direction-dependent modulation of parietal connectivity using a new brain functional connectivity measure

The posterior Parietal Cortex (PPC) of human and nonhuman primates plays a vital role in motor planning. However, EEG functional connectivity correlates within PPC associated with motor intentions are less investigated in the literature. In this study, we investigate whether parietal EEG exhibits direction-dependent modulation of functional connectivity, during bidirectional hand movement imagination in right and left directions. Further, the utility of parietal connectivity modulation patterns, in decoding the directions of imagined hand movement is also evaluated. Imagined movement directions of the dominant hand are decoded using connectivity features derived from parietal EEG. A new brain functional connectivity measure called Cumulative Phase Lag is proposed to evaluate the functional connectivity within the right and left hemispheres of the posterior parietal cortex. Parietal connectivity features are derived from twenty-three EEG subbands from both hemispheres. Further, hemispherical asymmetry is exploited to identify the hemisphere with dominant directive discriminability. Connectivity features of the selected hemisphere are used to identify the most discriminative subband and selected features of the discriminative subband are used to classify the directions of imagined hand movement. The proposed algorithm employing subject-specific connectivity features yielded an average right vs left-hand motor imagery direction decoding accuracy of 79.67 % among 15 healthy subjects. The study results revealed that connectivity patterns in the posterior parietal cortex exhibited direction-dependent variability, suggesting a direction-dependent modulation of connectivity within the posterior parietal cortex. The results suggest the use of the posterior parietal cortex as a potential source of control signals for neuro-prosthetic applications.