Oscillating Mindfully: Using Machine Learning to Characterize Systems-Level Electrophysiological Activity During Mindfulness Meditation
Year: 2022
Type: preprint
Abstract: There is rapid growth of neuroelectrophysiological study aspiring to uncover the “black box” of the mindful brain. Reliance on traditional data analysis methods hinders understanding of the complex, nonlinear, multi-dimensional and systemic nature of mindfulness functional neuroelectrophysiology. Thus, to reveal complex systemic neuroelectrophysiology of mindfulness meditation, we applied Machine Learning (ML) Extreme Gradient Boosting (XGBoost) classification algorithm, and four complementary feature importance methods, to extract systemic EEG features characterizing mindful states from EEG recorded during Focused Attention meditation and a control mind-wandering state among 26 experienced meditators. XGBoost classified mindful vs. mind-wandering states with 83% accuracy, AUC-ROC of 79% and F1 score of 74%. Feature importance techniques identified 10 EEG features associated with increased power and coherence of high frequency oscillations during mindfulness meditation. Findings delineate the complex systemic oscillatory activity characterizing mindfulness meditation.
Language: English
Subfield: Clinical Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4297664338