P 197. Frequency-specific directed connectivity and information flow in large-scale synchrony networks in the human brain-A TMS-EEG manipulative approach

Abstract

Synchronous neural oscillations are important in mediating information processing in the brain. To our knowledge, no study, however, has succeeded in analyzing directed connectivity and information flow in large-scale synchrony networks in humans. We therefore propose a novel manipulative approach for estimating frequency-specific network connectivity and information flow. We analyzed how single-shot TMS-modulated phase dynamics of ongoing oscillations at one cortical area propagates to the rest of the brain using phase resetting and amplitude measures in combination with an information flow measure, namely, transfer entropy. Using TMS-compatible electroencephalography (EEG) amplifiers, we recorded TMS-modulated ongoing brain activity while participants (N = 26 in total) sit on a chair with their eyes closed or eyes open fixating a gray cross. We targeted single-pulse TMS to the left primary motor cortex (95% motor threshold) or the visual cortex (95% phosphene threshold). First, we computed the instantaneous phase and amplitude of the filtered EEG signals. Next we calculated the phase locking values across trials as a measure for TMS-induced phase resetting at each electrode at different frequencies. We analyzed spatial propagation of TMS-induced phase resetting and amplitude changes. Finally, we estimated directed information flow between EEG phase signals and amplitude signals by transfer entropy ( Schreiber, 2000 ). We found prominent phase resetting of ongoing oscillations elicited by TMS. The phase resetting propagated from the stimulated area to the rest of the brain at various frequency bands. The spatio-temporal patterns of propagation of phase resetting were frequency-specific and state-dependent ( Fig. 1 ). Specifically we found more prominent global propagation of phase resetting in the eyes-open condition than in the eyes-closed condition most prominently at the 10 Hz alpha band in both motor and visual area targeted experiments. Amplitude changes were also frequency-specific and state-dependent. Significant increase of directed information flow from the stimulated area to the rest of the brain estimated by transfer entropy was also observed in the phase dynamics in all experimental conditions. We propose that this manipulative approach by TMS/EEG in combination with the transfer entropy method is good for estimating directed connectivity and information flow in the cortical and/or thalamocortical synchrony networks in the intact human brain. Since propagation of TMS-induced phase resetting is a robust and causal phenomenon, we can adjust and evaluate parameters of causality measures such as transfer entropy and Granger causality, which can indirectly assess directed connectivity and information flow from time series data. The results indicate that alpha synchrony networks are more globally connected in eyes-open condition than in eyes-closed condition, which is not obvious from indirect methods such as phase synchronization analyses of ongoing EEG activity, presumably associated with differences in gating of visual information processing.