EEG-Based Analysis of Mindfulness Meditation: Neural Modulation of Smoking Urges Using Empirical Mode Decomposition and Machine Learning

Smoking poses significant health risks and is strongly associated with addiction and relapse due to impaired response inhibition. Mindfulness meditation has shown promise in mitigating smoking urges, yet the underlying neural mechanisms remain poorly understood. This study investigates the effects of mindfulness meditation on smokers using electroencephalography (EEG), a non-invasive tool for analyzing brain activity. EEG data were collected from 10 male smokers using a 32-electrode system with a sampling frequency of 1024 Hz. Participants abstained from smoking, caffeine, and medications for five hours before the experiment to minimize external influences. Preprocessing involved noise reduction using a Butterworth 4th-order bandpass filter, isolating five frequency bands: delta (0–4 Hz), theta (4–8 Hz), alpha (8–16 Hz), beta (16–32 Hz), and gamma (32–64 Hz). Empirical Mode Decomposition (EMD) was employed to extract key features—mean, kurtosis, skewness, and entropy—from the first two Intrinsic Mode Functions (IMFs), which capture significant non-linear and non-stationary neural patterns. Feature selection using the Chi-Square test identified 20 critical features, primarily from delta and theta bands, associated with emotional regulation and cognitive control. Classification models, including Support Vector Machine (SVM) and Ensemble methods, achieved the highest training accuracy of 80% and a testing accuracy of 77.8%, effectively distinguishing between high and low smoking urges. Post-meditation EEG analysis revealed increased delta and theta activity, reduced beta activity, and behavioral improvements such as enhanced relaxation, reduced stress, and improved logical reasoning among participants. These findings underscore the potential of EEG-based analyses in understanding the neurobiological impact of mindfulness on smoking behavior. Future research should include larger sample sizes, longer meditation durations, and additional classifiers to enhance predictive accuracy and deepen insights into mindfulness as an intervention for smoking cessation.