Network analysis of meditative states in highly skilled meditators using EEG and horizontal visibility graphs.

The benefits of meditation are increasingly recognized, and some forms are now used for clinical intervention. However, the electrophysiological correlates of meditative states are not yet well understood, and the limited predictive accuracy of known markers of meditation suggest that not all information relevant to meditation has been captured by previous work.Here, we convert electroencephalography (EEG) time series into scale-free networks using horizontal visibility graphs (HVGs), which are well-suited to distinguishing deterministic dynamical systems from stochastic systems, allowing them to model novel aspects of cortical oscillatory activity. Based on HVGs, we introduce and evaluate a general class of predictors, which can be used to augment existing features in contemplative neuroscience, and exhibit high predictive power for several types of meditation.We show the statistical significance of these network predictors - and their increased performance compared to popular spectral and non-linear features such as complexity or entropy - on data from highly skilled meditators, in a continuous setting applicable to real-time analysis and applications such as neurofeedback.