摘要
目的 分析84例脑梗死后继发癫痫的危险因素,并建立XGBoost模型,分析该模型对继发癫痫的预测价值。方法 选取2019年6月至2022年8月郑州大学第一附属医院收治的84例脑梗死后继发癫痫患者纳入研究组,同时选取同期84例脑梗死后未继发癫痫患者纳入对照组。统计研究组癫痫发作情况。通过单因素、多因素logistic回归分析脑梗死后继发癫痫的影响因素。建立XGBoost模型,并采用受试者工作特征曲线、校准曲线分别评价该模型对继发癫痫的预测效果。结果 研究组中早发性癫痫51例、迟发性癫痫33例,单次发作56例、反复发作28例;较大梗死面积、颈动脉循环系统、美国国立卫生研究院卒中量表(NIHSS)评分增加、机械取栓、出血转化及血清同型半胱氨酸(Hcy)、神经元特异性烯醇化酶(NSE)、血清可溶性白细胞介素-2受体(sIL-2R)、脂蛋白相关磷脂酶A2(Lp-PLA2)水平升高是继发癫痫的危险因素(P<0.05);XGBoost模型依据不同危险因素进行重要度评分排序,依次分别为梗死面积、Lp-PLA2水平、NIHSS评分、sIL-2R水平、出血转化、NSE水平、机械取栓、Hcy水平、受累动脉系统;XGBoost模型预测脑梗死后继发癫痫的曲线下面积为0.826,且Hosmer-Lemeshow拟合优度检验(χ^(2)=7.015,P=0.541)。结论 梗死面积增大、颈动脉循环系统、NIHSS评分增加、机械取栓、出血转化及血清Hcy、NSE、sIL-2R、Lp-PLA2水平升高是脑梗死后继发癫痫的危险因素,建立XGBoost模型有助于筛选高危人群,该模型对继发癫痫具有一定预测价值,有助于指导临床防治脑梗死后继发癫痫。
Objective To analyze the risk factors of secondary epilepsy after cerebral infarction in 84 patients,and establish the XGBoost model to analyze the predictive value of this model for secondary epilepsy.Methods A total of 84 patients with secondary epilepsy after cerebral infarction from June 2019 to August 2022 were selected as the study group,and 84 patients with no secondary epilepsy after cerebral infarction during the same period were selected as the control group.Seizures were measured in the study group.Univariate and multivariate logistic regression analysis of the influencing factors of secondary epilepsy after cerebral infarction.The prediction effect of XGBoost model on secondary epilepsy was evaluated by receiver operating characteristic curve and calibration curve.Results In the study group,there were 51 cases of early onset epilepsy,33 cases of delayed onset epilepsy,56 cases of single seizure,28 cases of repeated seizure.Large infarct size,carotid circulatory system,increased National Institute of Health stroke scale(NIHSS)score,mechanical thrombectomy,bleeding transformation and elevated serum levels of homocysteine(Hcy)、neuron specific enolase(NSE),soluble interleukin-2 receptor(sIL-2R),lipoprotein associated phospholipase A2(Lp-PLA2)were the risk factors for secondary epilepsy(P<0.05).XGBoost model ranked the importance scores of different risk factors,which were infarct size,Lp-PLA2 level,NIHSS score,sIL-2R level,hemorrhage transformation,NSE level,mechanical thromboltomy,Hcy level,and involved arterial system in order.XGBoost model predicted an area under curve of 0.826 for secondary epilepsy after cerebral infarction,and Hosmer-Lemeshow goodness of fit test(χ^(2)=7.015,P=0.541).Conclusion The increase of infarct size,carotid circulation system,NIHSS score,mechanical thrombolysis,hemorrhage transformation and serum Hcy,NSE,sIL-2R and Lp-PLA2 levels are risk factors for secondary epilepsy after cerebral infarction.The establishment of XGBoost model is helpful for screening high-risk groups,and the model has certain predictive value for secondary epilepsy.It is helpful to guide clinical prevention and treatment of secondary epilepsy after cerebral infarction.
作者
郑婷婷
谢娜
袁萍
李玉生
ZHENG Tingting;XIE Na;YUAN Ping;LI Yusheng(Department of Neurology,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450000,China)
出处
《河南医学研究》
CAS
2024年第11期1990-1994,共5页
Henan Medical Research
基金
2021年河南省医学科技攻关省部共建重点项目(SBGJ202101016)。
