彩色密度谱阵列在昏迷患儿早期预后评估和识别痫样发作中的应用

发布时间：2018-05-19 06:50

本文选题：PICU + 昏迷　；参考：《吉林大学》2017年硕士论文

【摘要】：背景及目的:判明昏迷患儿预后和识别癫痫发作都是PICU经常面临的临床难题。本研究评估定量脑电图CDSA对PICU昏迷患儿的预后价值及其在癫痫发作监测方面的作用,以期建立简单直观的PICU床旁脑功能CDSA监测方法。方法:1)收集我院PICU住院且符合入组标准的昏迷患儿42例,在入院3天内首先进行Glasgow(GCS)评分,然后脑电监测仪记录CDSA不少于16h,随访3个月为终点。依据儿童脑功能分类量表,将正常、轻度残疾、中度残疾归为预后良好组(n=21),将严重残疾、昏迷或植物状态、脑死亡、临床死亡归为预后不良组(n=21)。回放CDSA和原始脑电图数据,从背景活动、醒睡周期、睡眠分期、药物性快波、双侧大脑半球对称性等方面评估,并通过建立受试者工作特征曲线(Receiver Operating Characteristic,ROC)与GCS评分比较,对比两种方法对预后的评估效力。2)选取小儿神经科脑电室日常27次视频脑电图记录,经傅里叶转化为4h/屏的CDSA图形,选取其中27张,标记出共256次区别于背景的图形改变,其中109处为发作事件,147处为对照。选取6名受试者(脑电图医生1名、脑电图技师1名、分别对照PICU医生2名、PICU护士2名),经2h包括CDSA基本原理、发作期图形、识别伪差等方面的培训后,在不接触原始脑电图数据的情况下,分别对27张CDSA图形中的256次图形改变进行判读,标记痫样发作事件,分析上述受试者判断发作性事件的一致性、漏判率、误判率及发生错误原因。结果:1)昏迷患者共42例,男女各21例,平均年龄89.81±46.39月。预后良好组21例,预后不良组21例,预后不良率50%。不同预后之间性别、年龄和病因构成无统计学差异(P0.05)。研究表明,与预后相关的因素包括醒睡分期、睡眠分期、药物性快波和CDSA分型。Logistic回归分析显示,醒睡周期是影响昏迷预后的独立危险因素。以预后不良为金标准,醒睡周期、GCS评分构建ROC曲线,结果表明,醒睡周期对判断昏迷患儿预后不良的效力优于GCS评分。2)6受试者应用CDSA图形识别癫痫发作的正确次数平均为82.6±3.5,敏感性为75.7%左右,其中癫痫持续状态均被正确识别。6人平均遗漏率约为24.1%,平均遗漏次数为26.3次,应用CDSA图形判读发作性事件的特异性达80%。经秩合检验比较,受试者间的敏感性和特异性无统计学差异(P0.05)。结论:1)CDSA图形中背景活动的可变化型、醒睡周期、睡眠分期、对药物具有反应性与昏迷患儿预后良好密切相关;无光谱型、单一慢型与昏迷患儿死亡密切相关。无醒睡周期是昏迷患儿预后不良的独立危险因素。醒睡周期对预后的评估效力优于GCS评分。2)CDSA识别癫痫发作具有中等敏感性,较低的误判率和漏判率。PICU医护人员经短期培训后即可达到较为理想的判读能力,PICU使用CDSA监测昏迷患儿脑功能具有应用前景和可行性。
[Abstract]:Background and objective: to identify the prognosis and identify epileptic seizures in children with coma are often the clinical problems faced by PICU. The purpose of this study was to evaluate the prognostic value of quantitative electroencephalogram (CDSA) in children with PICU coma and its role in monitoring epileptic seizures in order to establish a simple and intuitive CDSA monitoring method for PICU bedside brain function. Methods 42 cases of coma children who were admitted to our hospital with PICU were collected. Glasgow GCSs were scored within 3 days of admission, then CDSA was recorded by EEG monitor for no less than 16 hours, and followed up for 3 months as the end point. According to the children's brain function scale, normal, mild and moderate disability were classified as good prognosis group, severe disability, coma or vegetative state, brain death and clinical death were classified as poor prognosis group. CDSA and original EEG data were replayed and evaluated from background activity, sleep cycle, sleep stage, drug fast wave, bilateral cerebral hemispheres symmetry, and compared with GCS score by establishing receiver Operating characteristic roc. To compare the effectiveness of two methods in evaluating prognosis. 2) 27 times of daily video EEG records were recorded in the electroencephalogram (EEG) of the neurologic department of children, which were transformed into CDSA images of 4h/ screen by Fourier transform, and 27 of them were selected to mark out a total of 256-times image changes different from the background. Of these, 109 were seizures and 147 were controls. Six subjects (1 EEG doctor and 1 EEG technician) were selected and compared with 2 nurses in PICU. After 2 hours' training including basic principles of CDSA, pattern of attack period, identification of pseudo-error, etc. In the absence of contact with the original EEG data, the changes of 27 CDSA images were interpreted, and the epileptiform seizure events were labeled. The consistency of judging the paroxysmal events and the rate of missing judgment were analyzed. Error rate and cause of error. Results there were 42 comatose patients, 21 males and 21 males, with an average age of 89.81 卤46.39 months. There were 21 cases in good prognosis group and 21 cases in poor prognosis group, the rate of poor prognosis was 50%. There was no significant difference in sex, age and etiological composition among different prognosis (P 0.05). The results showed that the prognostic factors included sleep stage, drug fast wave and CDSA classification. Logistic regression analysis showed that sleep cycle was an independent risk factor for coma prognosis. With poor prognosis as the gold standard, the ROC curve was constructed by waking and sleeping cycle. The results showed that, The effectiveness of waking and sleeping cycle in judging the poor prognosis of coma children was better than that in the subjects with GCS score. 2The correct times of using CDSA pattern to identify epileptic seizures were 82.6 卤3.5 on average, and the sensitivity was about 75.7%. The average rate of omission was about 24.1and the average number of omissions was 26.3. The specificity of interpreting paroxysmal events with CDSA pattern was 80%. There was no significant difference in sensitivity and specificity among subjects by rank test (P 0.05). Conclusion the changes of background activity, sleep cycle, sleep stage, drug reactivity and prognosis of coma children are closely related to the changes of background activity in the CDSA pattern of 1: 1 CDSA, while the absence of spectrum type and the single slow type are closely related to the death of the comatose children. Unawake sleep cycle is an independent risk factor for poor prognosis in coma children. The effectiveness of sleep cycle in evaluating prognosis was better than that in GCS score. 2CDSA was of moderate sensitivity in the identification of epileptic seizures. The lower misjudgment rate and missed judgment rate. After short-term training, the medical staff in PICU can achieve a more ideal interpretation ability. The application prospect and feasibility of using CDSA to monitor brain function of coma children in PICU.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R720.597

【相似文献】