This study investigates the impact of OM chanting on human brain activity through the analysis of EEG signals. Utilizing a dataset of over 10,000 EEG recordings, we extracted features such as skewness, variance, kurtosis, and Shannon entropy. Machine learning models, including Support Vector Machines (SVM) and Random Forests, were employed to classify pre- and post-meditation states. The Random Forest model achieved an accuracy of 84.2%, outperforming the SVM model. SMOTE was applied to handle class imbalance, further enhancing accuracy to 89.1%. These findings provide quantitative insights into the neurophysiological effects of OM chanting, contributing to the scientific understanding of sound-based meditation practices.
Goyal et al. (Mon,) studied this question.