Brain Functional Network And Microstate Analysis During The Emergence Of Propofo-Iinduced Anesthesia

Reversible inhibition of consciousness through anesthetics provides the conditions for surgical treatment and other invasive medical examinations.A complete anesthesia process includes induction,maintenance and emergence.Delayed anesthesia emergence time is closely related to the incidence of post-operative dizziness,nausea,confusion,delirium and other diseases.However,most of the existing studies focus on the analysis of the relationship between the functional suppression of different nerve nuclei and the loss of consciousness during the induction of anesthesia,and few studies analyze the recovery process of the nervous system when the level of anesthetic concentration reduce gradually.Previous studies showed that the anesthesia emergence time varies greatly among different patients,which cannot be removed by pharmacodynamic and pharmacokinetic models,indicating that the nervous system plays an important role in the emergence from anesthesia.Therefore,it is of great significance to study the relationship between the anesthesia emergence time and the nervous system of individuals for the post-operative management and the prevention of anesthesia complications.In this paper,the relationship between the function of the nervous system and the anesthesia emergence time was explored from three aspects with the using of EEG signals obtained from the propofol anesthesia.First,EEG signals were recorded before and during propofol infusion.We calculated the energy distribution of each frequency band to obtain the features of EEG spectrum.In the meantime,pre-operative and intra-operative brain functional networks were constructed to obtain the graph measures.The relationship between calculated parameters and individual anesthesia emergence time were investigated by using linear regression analysis.Our results showed that there was no significant correlation between the level of intra-operative EEG suppression and the anesthesia emergence time,while the pre-operative brain functional network measures and PSD value were significantly correlated with the anesthesia emergence time.Secondly,in order to further explore the emergence from anesthesia,we used EEG microstates analysis method to study the transient state of brain neuron activity before and during operation,which has high temporal resolution.At the same time,the optimal cluster number of EEG microstates that can represent the activation of different brain networks during anesthesia was analyzed,and the trend of EEG microstates parameters with anesthesia emergence time was explored.Our results showed that the optimal cluster number of EEG microstates is 7,and the parameters of M4,M5 and M6 will change significantly with the anesthesia emergence time.Thirdly,the EEG microstate analysis method was used to calculate the parameters of preoperative and intra-operative EEG microstates sequence as the training features.LASSO was used to screen the features and train the prediction model,and the prediction effect of the screened features on anesthesia emergence time was verified by SVR.The results showed that the microstates parameters calculated by the pre-operative EEG had a better prediction effect on the anesthesia emergence time than the intra-operative microstate parameters,and the mean absolute percentage error of the prediction model was about 7%.