EEG alpha/beta features as a biomarker for quantifying pain in patients with lumbar disc herniation

Rumei Li ¹ Wanqi Shao ² Shumei Zhao ^2* Lingli Wang ^2* Chao Yu ^1* Lanying Liu ^2* Kuiying Yin ^1*

• ¹ Nanjing Research Institute of Electronic Technology, Nanjing, China
• ² Affiliated Hospital of Nanjing University of Chinese Medicine, Nanjing, Liaoning Province, China

Introduction: An objective and precise pain evaluation is of significant clinical value, and electroencephalography as a non-invasive physiological signal has been demonstrated to correlate with subjective pain perception. This study aimed to analyze the EEG changes in patients with lumbar disc herniation (LDH) under traditional Chinese medicine small needle knife and to further explore the feasibility of EEG as an indicator of pain assessment in patients with LDH.Methods: This study conducted resting-state electroencephalography on 20 patients with LDH before and after treatment and on 20 healthy controls, respectively. Following the spectral analysis of the EEG signals with continuous wavelet transform, power ratios were extracted for four frequency bands (θ, α, β and γ). Significance tests were conducted within the LDH group and between the LDH and healthy controls, as well as correlation analyses of EEG characteristics with pain scales in four regions of interest.Results: A significant reduction in subjective pain intensity was observed after small needle knife, with a 32.86% and 38.41% reduction in the Visual Analogue Scale (VAS) and modified Japanese Orthopaedic Association (mJOA) scores, respectively. Alpha accounted for a significantly higher of the four regions, while theta in the frontal, occipital and beta in the central were significantly lower. HC had fewer EEG oscillations in the theta band compared to LDH. The constructed alpha/beta features demonstrated a significant negative correlation with VAS in the frontal (R = -0.361, P = 0.022) and parietal (R = -0.341, P = 0.031), as well as with mJOA in the frontal (R=-0.416, P=0.007), central (R=-0.438, P=0.004), and parietal (R=-0.390, P=0.013) regions.EEG power ratios showed significantly different results in LDH groups, and between patients and HC. The alpha/beta features of the frontal and parietal constructed in this study showed correlations with subjective pain scores and might serve as a biomarker of pain status in the short term in LDH.