| General anesthesia is a drug-induced reversible state that along with continuously stable vital signs,and it is an important guarantee for safe operation.It is always desirable to achieve and maintain the depth of anesthesia that suitable for surgery with minimal anesthetic dosage.The main technologies to achieve this goal are precise anesthetic pharmacokinetic pharmacodynamic(PK-PD)model and effective mean of assessing the depth of anesthesia.Propofol is one of the most widely used anesthetics in the world,and there have been many PK-PD models for propofol.Moreover,with the development of brain functional measurement techniques and analytical algorithms in recent years,there have been some brain function related features that are proposed to monitor the depth of anesthesia.However,the existing propofol PK-PD models and anesthesia depth monitoring means still have some problems need to be solved.Firstly,the existing propofol PK-PD model mainly considers the effect of factors such as weight,age,and gender on the individual susceptibility to propofol.Recent studies suggested that the pre-anesthesia brain function status may also be able to affect the individual susceptibility to propofol.However,the existing PK-PD model did not take this factor into consideration.Secondly,it has been shown that there exist differences in the anesthetic drug concentration at the same anesthesia level during the process of anesthesia induction and emergence,which is defined as hysteresis during anesthesia.The current anesthetic PK-PD models consider this hysteresis as a pharmacokinetic delay and eliminate it by optimizing the model parameter.However,recent animal studies suggested that the observed hysteresis may be an intrinsic characteristic of the neural system.Therefore,it is controversial to eliminate the observed hysteresis in the PK-PD model.Thirdly,numerous studies mainly employ a static approach to explore changes in the brain network during anesthesia.The implicit assumption of this method is that brain activity is stable throughout the duration of data recording.However,recent studies showed that the non-stationary characteristics of brain functional activity still exists at anesthetic-induced unconsciousness.Hence,investigating the dynamic changes of brain functional network during anesthesia is the key to explore reliable anesthesia depth monitoring means.Finally,the EEG features that are available for monitoring the depth of anesthesia are mainly obtained by comparing the differences between wakefulness and unconsciousness.However,accurate anesthesia depth monitoring means not only need to distinguish wakefulness and unconscious state,but also need to monitor continuous changes of conscious level during anesthesia,i.e.the fluctuation of conscious level.Therefore,exploring the EEG features that can detect fluctuations of conscious level is of great value in optimizing the monitoring methods of anesthesia depth.In this study,we carried out a series of work to address these issues.The main results and innovations are as follows:1.We explored the relationship between the pre-anesthesia brain functional status and individual susceptibility to propofol.To do so,we collected the resting state EEG data prior to anesthesia,and then used the sliding window analysis to obtain pre-anaesthesia strength and fluctuation of frontal-parietal functional connectivity,which were used to measure the pre-anesthesia brain functional status of subjects.The individual susceptibility to propofol was quantified by the time duration of the start of propofol infusion to achieve the same anesthesia level.The results showed that the individual susceptibility to propofol was correlated with the estimated pre-anesthesia dynamic frontal-parietal functional connectivity,and the results were verified on three independent EEG data sets.These findings indicate that the individual susceptibility to propofol is influenced by the pre-anesthesia brain functional status.Therefore,when designing PK-PD models of propofol,the patient's preoperative brain functional status should also be considered into the model.2.We investigated wether the hysteresis observed during anesthesia is an intrinsic characteristic of the neural system.The bispectral index(BIS)and brain functional network measures were employed as the index for anesthetic effect-site concentration and anesthetic effect,respectively.We found that there existed significant differences in the brain functional network measures between the process of anesthesia induction and emergence.Therefore,it can be inferred that the hysteresis observed during anesthesia was an intrinsic characteristic of the brain.Our results suggest that the hysteresis should be taken into consideration when designing PK-PD models for the anesthetics such as propofol,and the hysteresis index proposed in the present study can be used as a measure to quantify the magnitude of individual hysteresis.3.We investigated the changes of dynamic brain functional network from awaken state to the anesthesia level suitable for surgery.To do so,we collected EEG data at wakefulness,light and deep anesthesia,and then used EEG source imaging analysis to locate the cortical activity.Then,k means clustering analysis was employed to identify the common dynamic brain functional network patterns.We also analyzed the transition behavior of these reoccurring network patterns at different conscious states.The results showed that the functional network pattern which is supported by the fundamental anatomical structure of the brain was preserved during anesthesia,which might explain why the subjects can maintain physiologically stable during anesthesia.Furthermore,the results showed that the frequency of occurrence of each brain network pattern was associated with the level of anesthesia,which provide the possibility of using brain network features to monitor the depth of anesthesia.4.We explored the performance of existing consciousness-related EEG features for detecting the fluctuation of conscious level.The behavioral test was employed to assess the level of consciousness of the subjects during anesthesia.Totally 110 EEG features,including the power spectrum,Synch Fast Slow,permutation entropy,topographical proportion of alpha power,functional connectivity and graph measures of brain functional network were calculated based on EEG data that collected during propofol anesthesia.Then,we used classification analysis to evaluate the performance of these features in detecting the fluctuation of conscious level.The results showed that the existing consciousness-related EEG features had a high accuracy rate of 93.5% for identifying wakefulness and unconsciousness,but they were not sensitive enough to detect the fluctuation of conscious level.The features including theta band power,functional connectivity and graph theory measures and the proportion of alpha power in the posterior region of the brain were more sensitive to detect fluctuation in conscious level as compared to other features.In conclusion,we conducted a series of studies with the aim of improving PK-PD models of propofol and exploring the effective way to assess the anesthesia depth based on EEG that are collected during wakefulness and propofol-induced anesthesia.These results not only help us better understand the mechanism of propofol-induced unconsciousness,but also provide valuable evidences for better anesthesia management in clinical settings. |
PDF Full Text Request