基于临床数据和HR-MRI相关参数构建脑梗死发生的预测模型OACSTPCD

Objective To construct a prediction model for the occurrence of cerebral infarction(ACI)due to intracranial atherosclerosis(ICAS)plaque formation based on clinical data and vascular wall imaging.Methods Totally 204 patients with ICAS plaque formation in the Langfang People's Hospital from February 2018 to June 2022 were selected and divided into observation group(80 cases)and control group(124 cases)according to whether ACI occurred.The clinical data and HR-MRI images parameters of the two groups were collected and compared.LASSO-Logistic regression was used to analyze the risk factors for ACI,and Nomogram prediction model was drawn.Results The proportion of hypertension,diabetes mellitus,D-D,UA and Hcy in observation group were higher than those in control group(P<0.05).Plaque hemorrhage,plaque load,plaque volume,remodeling index,eccentricity index,plaque enhancement rate at the narrowest layer and overall plaque enhancement rate in observation group were higher than those in control group(P<0.05).Nine factors were initially screened by LASSO,including hypertension,diabetes,D-D,Hcy,plaque load,remodeling index,eccentricity index,plaque enhancement rate at the narrowest level,and plaque enhancement rate as a whole.Logistic regression analysis showed that the above nine factors were independent risk factors for ACI caused by ICAS plaque formation(P<0.05).The Nomogram prediction model of ACI induced by ICAS plaque formation was constructed,and the C-index of the model was 0.944.ROC curve analysis showed that the AUC for predicting ACI caused by ICAS plaque formation was 0.913(0.857-0.972).According to the calibration chart analysis,the calibration degree of the model was 0.888.Conclusion The prediction model of ACI induced by ICAS plaque formation based on clinical data and HR-MRI related parameters is feasible and reliable,which can provide guidance information for clinical prevention and treatment.