Meditation Alters Cognitive Load: EEG Fractal-Based Classification and HRV Geometric Assessment

Cognitive load during demanding tasks can be measured through brain and heart signals. After eight weeks of Hare Krishna Mantra (HKM) meditation, participants showed smaller changes in EEG complexity and heart rate variability between rest and task states, indicating reduced cognitive load and more stable brain-heart dynamics. Machine learning classifiers distinguished rest from task periods with 80-81% accuracy before meditation, but only 61-62% after, confirming that meditation made these states less distinguishable.

Study at a glance

Abstract

Objective measures of cognitive load are essential for understanding how the brain and body respond to demanding tasks and how meditation modulates these responses. This study examines nonlinear complexity features from electroencephalography (EEG) and heart rate variability (HRV) for distinguishing rest from task states and assessing meditation effects during cognitive tasks. Cognitive load was quantified using the Hurst exponent (HE) and Higuchi fractal dimension (HFD) from EEG, capturing temporal correlations and signal irregularity, and the SD2/SD1 ratio from HRV as an index of sympathovagal balance. Classification between rest and task periods was performed using Support Vector Machine (SVM) and Random Forest (RF). Task performance showed decreased HE, increased HFD, and higher SD2/SD1, indicating elevated load and sympathetic dominance. After eight weeks of Hare Krishna Mantra (HKM) meditation, smaller variations in HE, HFD, and SD 2/SD 1 were observed, and SVM accuracy decreased from 80.1% to 61.0% while R F accuracy dropped from 81.4% to 62.0%, confirming reduced separability between rest and task states. These findings demonstrate that HKM meditation lowers cognitive load and stabilizes brain-heart dynamics, supporting EEG and HRV features as effective multimodal indicators of mental effort.