Design and Implementation of Electroacupuncture: A Study of Prefrontal EEG Characteristics Under taVNS
Lixian Zhu, Yanan Zhao, Xiaokun Jin, Fuze Tian, Jingxin Liu, Ran Cai, Qunxi Dong, Peijing Rong, Bin Hu
Abstract
Transcutaneous auricular vagus nerve stimulation (taVNS), as a method for mimicking VNS, has been proven effective in the treatment of psychiatric disorders. However, the underlying mechanism through which taVNS mimics VNS remains elusive. Moreover, the parameters of taVNS are singularly fixed and open loop in previous work, which is difficult to apply to all users as individual differences are inevitable. Since electroencephalogram (EEG) is one of the important biomarkers of neural activity, this study aims to develop a closed-loop system for personalized interventions in emotion regulation by integrating taVNS with EEG feedback. We first design a taVNS system based on EEG signal feedback and verify the performance metrics of the system. Second, we design experimental paradigms to explore the changes in EEG features under the taVNS. The experimental results show that the EEG characteristics differ between different taVNS frequencies (between 50 and 100 Hz). Moreover, we observe substantial distinctions between EEG characteristics during the taVNS state and the resting state, with pre-taVNS, taVNS, and post-taVNS exhibiting notable differences. Specifically, the power spectral density (PSD) in the taVNS state is lower than in the resting state (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${p} \lt {0.05}$ </tex-math></inline-formula>), except for the beta band where the opposite trend is observed. Additionally, features such as Lempel-Ziv complexity (LZC) and Reyi entropy (REn) displayed a decreasing trend throughout the taVNS (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${p} \lt {0.05}$ </tex-math></inline-formula>). Furthermore, we employ hidden Markov models (HMMs) to reveal the heterogeneity of dynamic changes in the brain during taVNS, providing a mechanistic interpretation of taVNS.