基于定量脑电图与多指标特征构建急性重症脑出血早期功能恶化的预测模型  

Construction of a Predictive Model for Early Functional Deterioration in Acute Severe Cerebral Hemorrhage Based on Quantitative Electroencephalography and Multi-indicator Features

作　　者：于朝旭[1] 葛晶[2] 王迎东 张歆萍[1] 张晓娟 孙实安[1] YU Chao-xu;GE Jing;WANG Ying-dong;ZHANG Xin-ping;ZHANG Xiao-juan;SUN Shi-an(a.Department of Neurosurgery,Chengde Central Hospital,Chengde,Hebei 067000,China;Department of Neurodiagnosis,Chengde Central Hospital,Chengde,Hebei 067000,China)

机构地区：[1]承德市中心医院神经外科,河北承德067000 [2]承德市中心医院神经诊断科,河北承德067000

出　　处：《临床误诊误治》2023年第1期75-80,共6页Clinical Misdiagnosis & Mistherapy

基　　金：承德市科学技术局项目(202109A184)。

摘　　要：目的探究基于定量脑电图与多指标特征构建的预测模型对急性重症脑出血患者早期神经功能恶化(END)的预测价值。方法收集2019年10月—2021年10月152例急性重症脑出血临床资料,统计入院72 h内END发生情况,比较发生与未发生END患者一般资料及脑电双频谱指数(BIS),实验室指标[神经元特异性烯醇化酶(NSE)、D-二聚体(D-D)、凝溶胶蛋白(GSN)、Caspase切割的细胞角蛋白-18(CCCK-18)、轴突生长抑制因子-A(Nogo-A)、核因子-κB p65(NF-κB p65)、微小RNA-130a(miR-130a)、全身炎症反应指数(SIRI)],分析END发生的影响因素。基于BIS和其他影响因素构建Nomogram预测模型,采用决策曲线分析含与不含BIS的Nomogram预测模型的临床净获益。结果152例END发生率为28.29%(43/152)。血肿体积、入院美国国立卫生研究院卒中量表评分及血清NSE、D-D、CCCK-18、Nogo-A、NF-κB p65、SIRI、miR-130a水平为急性重症脑出血患者END发生的独立危险因素,入院格拉斯哥昏迷量表评分、BIS及血清GSN水平为急性重症脑出血患者END发生的保护因素(P<0.01)。基于BIS与多指标特征构建的Nomogram预测模型预测急性重症脑出血患者END发生的一致性指数为0.881(95%CI为0.831,0.940),具有良好的区分度;校准曲线显示该模型与实际观测结果有较好的一致性;决策曲线显示含BIS的模型预测值在0~0.9区间时,可提供附加临床获益。结论BIS是急性重症脑出血患者END发生的影响因素,基于BIS与多指标特征构建的Nomogram预测模型能提高对急性重症脑出血患者END发生的预测价值。Objective To explore the predictive value of the prediction model based on quantitative electroencephalogram(EEG)and multi-indicator characteristics for early neurological deterioration(END)in patients with acute severe cerebral hemorrhage.Methods The clinical data of 152 patients with acute severe cerebral hemorrhage from October 2019 to October 2021 were collected.The occurrence of END within 72 h after admission was statistically analyzed,and the general data and BI spectral index(BIS)of patients with and without END were compared.Laboratory indicators[neuron-specific enolase(NSE),D-dimer(D-D),gelsolin(GSN),Caspase-cleaved cytokeratin 18(CCCK-18),axon growth inhibitor A(Nogo-A),nuclear factor-κB p65(NF-κB)p65,micrornA-130A(miR-130a),systemic inflammatory response index(SIRI)]were compared to analyze the influencing factors of the occurrence of END.A Nomogram prediction model was constructed based on BIS and other influencing factors,and the net clinical benefits of this model with and without BIS was analyzed by using decision curves.Results The incidence of END in 152 cases was 28.29%(43/152).Hematoma volume,NIHSS score on admission and serum levels of NSE,D-D,CCCK-18,Nogo-A,NF-κB p65,SIRI and miR-130a were independent risk factors for END occurrence in patients with acute severe cerebral hemorrhage.GCS score on admission,BIS and serum GSN levels were protective factors for the occurrence of END in patients with severe acute cerebral hemorrhage(P<0.01).The consistency index of this Nomogram model based on BIS and multi-indicator characteristics of patients with acute severe cerebral hemorrhage was 0.881(95%CI:0.831,0.940),with good differentiation.The calibration curve showed that the model was in good agreement with the actual observation results.The decision curve provided additional clinical benefit when the BIS model prediction was in the range of 0 to 0.9.Conclusion BIS is an influencing factor for the occurrence of END in patients with acute severe cerebral hemorrhage.A Nomogram model based on BIS an

关 键 词：脑出血 早期神经功能恶化 脑电图 脑电双频谱指数 影响因素分析 预测模型 

分 类 号：R743.34[医药卫生—神经病学与精神病学]