基于LASSO-Logistic回归建立创伤性脑损伤患者继发认知障碍预测模型  

Prediction model of secondary cognitive dysfunction in patients with traumatic brain injury based on LASSO-Logistic regression

作　　者：查智健 耿锋 李光照[1] 杨非[1] ZHA Zhijian;GENG Feng;LI Guangzhao;YANG Fei(Hefei First People’s Hospital,Hefei 230092,China)

机构地区：[1]合肥市第一人民医院,安徽合肥230092

出　　处：《中国实用神经疾病杂志》2025年第3期316-320,共5页Chinese Journal of Practical Nervous Diseases

基　　金：安徽医科大学校科研基金项目(编号:2022xkj144)。

摘　　要：目的构建并验证基于LASSO-Logistic回归的创伤性脑损伤患者继发认知障碍预测模型。方法选取2021-01—2023-12合肥市第一人民医院收治的96例创伤性脑损伤患者为研究对象,根据是否继发认知障碍分为继发组和未继发组。采用LASSO-Logistic回归筛选创伤性脑损伤患者继发认知障碍的影响因素,并以此构建风险列线图模型,使用受试者工作特征(ROC)曲线验证其效能,利用Bootstrap法检验模型的准确性。结果96例创伤性脑损伤患者中继发认知障碍27例,发生率28.13%。继发组患者与未继发组患者年龄、糖尿病、高血压、格拉斯哥昏迷量表(GCS)评分、手术治疗、肿瘤坏死因子-α(TNF-α)、白细胞介素-6(IL-6)比较均有统计学差异(P<0.05)。LASSO-Logistic回归分析显示,年龄、糖尿病、高血压、GCS评分、手术治疗、TNF-α、IL-6可作为构建创伤性脑损伤患者继发认知障碍的预测因素。基于以上因素构建的风险列线图模型曲线下面积为0.805,特异度、敏感度分别为77.65%、80.97%,Hosmer-Lemeshow检验显示χ^(2)=3.664,P=0.729。结论基于LASSO-Logistic回归构建的创伤性脑损伤患者继发认知障碍预测模型具有较好的诊断效能,可为创伤性脑损伤患者的病情监测与管理提供参考。Objective To construct and verify the prediction model of secondary cognitive dysfunction in patients with traumatic brain injury based on LASSO-Logistic regression.Methods Ninety-six patients with traumatic brain injury admitted to Hefei First People’s Hospital from January 2021 to December 2023 were selected as the study objects,and were divided into the secondary group and the non-secondary group according to whether the patients had secondary cognitive dysfunction.LASSO-Logistic regression was used to screen the influencing factors for secondary cognitive dysfunction in TBI patients,and the risk Nomogram model was constructed based on it.The receiver operating characteristic(ROC)curves were used to verify its efficacy,and the accuracy of the model was tested by Bootstrap method.Results Among 96 patients with traumatic brain injury,27 had secondary cognitive dysfunction(28.13%).There were significant differences in age,diabetes,hypertension,GCS score,surgical treatment,TNF-αand IL-6 between the secondary group and the non-secondary group(P<0.05).LASSO-Logistic regression analysis showed that age,diabetes,hypertension,GCS score,surgical treatment,TNF-α,and IL-6 could be used as predictors of secondary cognitive impairment in TBI patients.Based on the above factors,the area under the curve of the risk Nomogram model was 0.805,and the specificity and sensitivity were 77.65%and 80.97%,respectively.Hosmer-Lemeshow test results show thatχ^(2)=3.664,P=0.729.Conclusion The prediction model of secondary cognitive dysfunction in TBI patients based on LASSO-Logistic regression has good diagnostic efficacy,and can provide reference for the monitoring and management of TBI patients.

关 键 词：创伤性脑损伤 LASSO-Logistic回归 认知障碍 预测模型 诊断效能 

分 类 号：R651.15[医药卫生—外科学]