Objective To investigate the changes of non-linear dynamics properties of EEG under different mental tasks,and the application of non-linear dynamic analysis for cognitive function research. Methods EEG was recorded in 30 healthy university students under four states: eyes closed,eyes open,mental arithmetic with eyes closed and graphic reasoning test with eyes open.Correlation dimension (D 2) and point-wise correlation dimension (PD 2) were calculated for all subjects.We also performed a surrogate data test in 5 randomized selected subjects. Results In all states tested,the results argued for the presence of non-linearity and fractal D 2 and PD 2.D 2 and PD 2 increased significantly during mental tasks as compared to the rest states (D 2 of above four states was 3.93,4.47,4.33 and 4.98 respectively).The results showed that values of D 2 and PD 2 fluctuated with time,which suggests an intermittent activation of the brain areas participated in the task.With non-linear dynamic analysis,there might be a working circuit from bilateral temporal lobes to the left frontal lobe and parental lobe under mental arithmetic task.Conclusions Through dynamic analysis based on D 2 and PD 2 topographic maps,we can see clearly the distribution of D 2 and PD 2 in the brain,the sequence and the degree of activation of the regions involved in the cognitive task under mental tasks.Dynamic and short-time non-linear analysis methods are more appropriate for the study of mental functions.Non-linear dynamic analysis might help us understand the working mechanism of brain during mental activities.

Objective To investigate the relationship between intraoperative cerebral oxygen saturation (rSO2) and postoperative cognitive dysfunction with near-infrared cerebral oximeter (INVOS 5100) in patients operated under inhalational combined intravenous anesthesia, and to determine the critical rSO2 value below which postoperative cognitive dysfunction may occur. Methods Sixty ASAⅠ-Ⅱ patients of both sexes were selected, aged 62-80yr, weighed 58-77kg, scheduled for elective abdominal surgery or surgery on the low limb. All the patients were divided into three groups according to their educational background: in group Ⅰ were the illiterate and uneducated patients (n=20);group Ⅱ the primarily educated patients (<6yr education) (n=20), and group Ⅱ the well educated patients (>6yr education) (n=20). Each group was further divided into isoflurane and sevoflurane subgroups (n=10 in each subgroup). All patients received no pre-medication. Anesthesia was induced with intravenous atropine 0.3mg, propofol 1.0-1.5mg kg-1, fentanyl 2-3μg*kg-1 and vecuronium 0.1-0.2mg*kg-1, and maintained with isoflurane or sevoflurane inhalation(0.9-1.1 MAC) supplemented with intermittent i.v. boluses of fentanyl, and recorded after entering room (baseline) (T0), after O2 inhalation (T1), after induction of anesthesia (T2), after skin incision (T3), during operation (T4), the end of surgery (T5), and awaking (T6). Mini-Mental State Examination (MMSE) was performed before anesthesia and 1, 4, 8, 12 and 24h after surgery. BP, HR, ECG, SpO2, PETCO2 and end-tidal concentration of inhalational anesthetics were continuously monitored during anesthesia. Results In all three groups rSO2 was significantly lower during operation (T4) and at the end of surgery (T5) than baseline (T0) (P<0.05). In all patients the MMSE scores at 1h after operation were significantly lower than the baseline value (P<0.05). The MMES scores in all patients significantly declined within 1-4h after surgery, and the cognitive function recovered at 4h after surgery in 85% patients. The critical values of rSO2 below which postoperative cognition dysfunction may occur were: 45 (group Ⅰ), 47 (group Ⅱ) and 49 (group Ⅲ) for isoflurane anesthesia subgroups;47 (group Ⅰ), 48 (group Ⅱ) and 50 (group Ⅲ) for sevoflurane subgroups. Conclusion The perioperative rSO2 should be maintained up to above 50% to reduce the incidence of postoperative cognitive dysfunction under inhalational combined intravenous anesthesia.

Objective To investigate the changes in non-linear dynamics properties of EEG and the application of real-time monitoring of non-linear indexes during the course of general anesthesia. Methods EEG was recorded in 65 patients undergoing operations. They were randomly divided into five groups: isoflurane, sevoflurane, desflurane (n=15, respectively), propofol 8mg/(kg?h) and propofol 10mg/(kg?h) (n=10, respectively). The EEG derived parameters correlation dimension (D 2 ) and approximate entropy (ApEn) non-linear indexes were calculated simultaneously during the whole operation, including the time of entering the operation room, anesthetic induction, intraoperation, recovery, and regaining consciousness. Results At the time of entering the operation room, their D 2 and ApEn nonlinear indexes appeared to be highest. Both the indexes decreased swiftly during anesthetic induction, and they dropped to lower values and leveled off during the intra-operation period. During recovery period, both of them rose gradually and returned to a high level in the post-operation awaking period. Conclusions Changes in the depth of anesthesia could be real-timely monitored and precisely measured with non-linear indexes of EEG. Non-linear dynamic analysis might provide us with more information on consciousness during general anesthesia

Objective To study the effect of inhalational anesthetics on brain cortices under inhalational and intravenous combined anesthesia. Methods 45 patients were randomly divided into isoflurane group (n=15), sevoflurane group (n=15) and desoflurane group (n=15). The narcosis was maintained with inhalational and intravenous combined anesthesia. The EEG non-linear parameters including approximate entropy (ApEn) and correlation dimension (D 2 ) were recorded during operation periods. BP, HR, and SpO 2 were monitored routinely. Results Comparing with that at entrance to the operating room, the EEG activities of frontal and temporal cortices after anesthesia were more suppressed than other cortices. Comparing with that at entrance to the operating room, ApEn and D 2 were significantly declined in the three experimental groups (P

Objective To study effect of propofol on different areas of brain cortex under total intravenous propofol and fentanyl anesthesia. Methods 20 patients were randomly divided into propofol 8mg/(kg?h) group (n=10) and 10mg/(kg?h) group (n=10). The anesthesia was maintained by total intravenous propofol and fentanyl anesthesia. The EEG non-linear parameters, i.e. approximate entropy (ApEn)and correlation dimension (D 2), were recorded during perioperative periods. BP, HR, SpO 2 were monitored routinely. Results Compared with that at entry of OR, the EEG activities of frontal, parietal, and temporal cortices after anesthesia were more inhibited than that of other areas in both groups. Also EEG non-linear parameters were lower significantly after anesthesia in both groups compared with those at entry of OR (P

0 05) and neither implicit nor explicit memory in other groups. D 2 and ApEn nonlinear indexes were higher while entering the operative room. Both indexes dropped to a relative lower value and leveled off in the intra operative period. In 5mg group, both indexes increased significantly during the memory test compared with intraoperative period (D 2 and ApEn nonlinear index were 4 07 and 3 30, 0 67 and 0 53 respectively, P

Objective The present study was undertaken to investigate the properties of nonlinear dynamics of EEG and the changes in depth of anesthesia with real-time approximate entropy (ApEn) and complexity (Cx) nonlinear indexes monitoring during anesthesia. Methods EEG was recorded in 65 in-patients. They were randomly divided into 4 groups: isoflurane, sevoflurane, desflurane (n=15, respectively), and propofol intravenous anesthesia (n=20) groups. The EEG derived parameters ApEn and Cx non-linear indexes were calculated simultaneously during the whole operation including rest state with eyes closed, anesthetic induction, intraoperation, recovery, post-operation awaking. Results ApEn and Cx nonlinear indexes remained the highest during rest state. Both of them kept decreasing during anesthetic induction. They dropped to a relative lower value and leveled off in the intra-operation period. Both of them rose gradually during recovery period and returned to a high level in the post-operation awaking period (correspondingly, ApEn: 0.87, 0.78, 0.55, 0.64 and 0.83. Cx: 0.58, 0.54, 0.38, 0.46 and 0.57). Conclusions With ApEn and Cx non-linear indexes, changes in depth of anesthesia from EEG signal could be real-timely monitored and precisely measured. Nonlinear dynamic analysis might provide us with more information about consciousness and cognition during general anesthesia.

Objective To analyze the bispectral index (BIS) and the EEG nonlinear index (including Correlation dimension, D2; Approximate entropy, ApEn; Complexity, Cx.) during alternating periods of consciousness and unconsciousness produced by target-controlled infusions (TCI) of propofol. Methods We studied twenty patients (ASA Ⅰ～Ⅱ grades) undergoing the elected leg operations under epidural anesthesia. With TCI consciousness of the patient was controlled by an increase or decrease of concentration of propofol in a range of 0.3～0.5?g/ml for four times. Every target plasma concentration of propofol lasted 12 minutes. BIS, D2, ApEn and Cx were recorded simultaneously during the periods of consciousness and unconsciousness every 3 minutes. Results During consciousness and unconsciousness, the respective mean values for the four measurements were: BIS, 80.2?6.2 and 67.3?7.9; D2, 3.45?0.18 and 3.01?0.16; ApEn, 0.84?0.05 and 0.71?0.06; Cx, 0.55?0.05 and 0.44?0.05. Determined threshold values with 100% specificity during the state of unconsciousness were: BIS, 51 (sensitivity3.8%); D2, 2.90 (sensitivity 30.3%); ApEn, 0.69 (sensitivity42.3%); Cx, 0.41 (sensitivity 25.5%). Conclusion BIS, D2, ApEn and Cx can all reflect the change in consciousness and unconsciousness produced by TCI of propofol. Our findings suggest that of the four EEG variables, ApEn was best in identifying the transition from unconsciousness to consciousness.

Objective: ①To study whether changes of cortical electrical activations in two hemispheres correlated with speech tasks could be observed with approximate entropy analysis of EEG; ②To dynamically observe the correlation of speech improvements with cortical electrical activations in two hemispheres during speech treatment. Method: Changes of cortical electrical activations were observed with EEG in a conductive aphasic 7、9 and 30 weeks after stroke. EEG was recorded in 3 different conditions: eyes closed, word repetition and non-word repetition. The EEG of a healthy man was recorded in 3 different conditions as control. Result: Compared with eye-closed condition, EEG approximate entropies(ApEn_s) of the healthy subject increased in T3, T5, C3, P3 and O1 for word and non-word repetition, and additional F7 and T6 for non-word repetition task. For the patient, ApEn of the first EEG increased in P3 for word repetition, and P4 for non-word repetition; ApEns of the second EEG increased in F3, F7, P3, C3, C4,T3, T5, O1, and decreased in F8, T4 and O2 in right hemisphere for word and non-word repetition; ApEns of the third EEG increased in T3, P3, F7, O1, O2 for word repetition,and T3, O2 for non-word repetition. The first aphasia assessment showed light impairment of semantic system and initial sound auditory identification and severely impairment of picture naming and word repetition. The semantic system and initial sound auditory identification improved significantly in the second assessment compared with the first; but the word repetition remained unchanged. The range of the activated regions in the left hemisphere decreased in the third EEG compared with the second EEG and the activated regions were less than the normal subject. Conclusion: The cortical electrical activities correlated with repetition tasks and the change of cortical electrical activities in different stages of language recovery could be observed with EEG nonlinear analysis.

Objective To investigate the influence of poly(ADP-ribose)polymerase inhibitor PJ34 on blood brain barrier(BBB)in rats with cere-bral ischemia-reperfusion injury. Methods Rat model of cerebral ischemia-reperfusion injury was established by the middle cerebral artery occlu-sion. A total of 135 SD rats were randomly divided into 3 groups:sham-operated group(sham group),ischemia-reperfusion group(IR group)and PJ34 group(PJ34 group). 45 animals in each group were then equally divided into subgroups and the rats were sacrificed at 6 h,24 h,48 h after re-perfusion,respectively. BBB permeability was evaluated by detection of extravasated Evans blue(EB). The expression of tumor necrosis factorα(TNF-α)and matrix metalloproteinase 9(MMP-9)activity were measured by immunohistochemistry and western blot at different time points. Re?sults Compared with sham group,the contents of EB and the expressions of TNF-αand MMP-9 in IR group were increased significantly(P<0.05). Compared with IR group,the contents of EB and the expressions of TNF-αand MMP-9 in PJ34 group were markedly decreased at the same time point(P<0.05). Conclusion The present study provided in vivo evidence that PARP inhibitor PJ34 can protect against cerebral ischemia re-perfusion injury,and the mechanism might be related to maintaining the stability of blood-brain barrier by suppressing the expression of TNF-αand MMP-9 in ischemic cortex.