Consumer-Grade Brain Measuring Sensor in People With Long-Term Kratom Consumption

Neurophysiological characteristics of long-term Kratom users have been challenged for identification due to the lack of evidence. Long-term and high Kratom consumption caused concern, particularly in older adults. Thus, the study aims to explore EEG biomarkers in long-term Kratom users (LKU) based on consumer-grade EEG systems. The fifty-two participants were collected EEG using MUSE portable system during resting-state to examine EEG biomarkers with the proposed features: theta/alpha ratio and power variance function (PVF) in theta and alpha bands. The statistical analysis was further carried out to test the existent difference between controls and LKU in various age ranges (<= 50 and >50 years of age) and different doses of Kratom consumption (low to high doses and very high dose). Subsequently, the statistical-based EEG biomarkers were extracted and performed classification among four classifiers (Random forest, Support vector machine, K-Nearest Neighbor, and Logistic regression. As a result, the TAR ratio was remarkably different between groups over 50 years of age. Furthermore, TAR and PVF in the alpha band were dominant in those who consumed Kratom at a very high dose and was classified well by support vector machine using the features combination (accuracy at 83.33% +/- 10.24, sensitivity at 90.00% +/- 10.00, specificity at 75.00% +/- 13.44). Our preliminary results concluded that the proposed EEG features were an important EEG biomarker for LKU with a large effect size. This finding led us to the promising aspect of applying machine learning-based EEG biomarkers to screen the overdose of Kratom consumption in the future.