Gamma-Power in a Fronto-Parietal Network Predicts Motor-Imagery Performance
While the neuro-physiological basis of motor imagery has been studied in great detail, little is known on the neural determinants of good or bad motor-imagery performance. In this talk, I present evidence based on EEG recordings in normal subjects that the baseline power of fronto-parietal gamma-oscillations (i.e., oscillations of the electromagnetic field of the brain roughly above 50 Hz) predicts the performance of subsequent motor-imagery on a trial-to-trial basis. Furthermore, our results suggest that the power of these gamma-oscillations is not modulated by the instruction to initiate motor imagery, but oscillates autonomously at predominantly very low frequencies. I analyze these observations in the framework of Causal Bayesian Networks, and argue that they provide support for a causal influence of a fronto-parietal resting-state network on motor-imagery performance.