Objective:By comparing 9 CT signs of complicated appendicitis(CA)and uncomplicated appendicitis(UCA),to screen the risk predictors of CA for the models'establishment.Methods:Clinical and CT data of 403 patients with acute appendicitis were collected.93 patients were in the CA group and 310 patients were in the UCA group.9 CT signs of two groups were observed and compared.403 patients were randomly divided into a training cohort(283 cases)and a validation cohort(120 cases)at a ratio of 7∶3.Univariate and multivariate logistic regression analysis and model construction were carried out on the training cohort.ROC curve and calibration curve were drawn to evaluate the discrimination and calibration degree of the model.Bootstrap method on the original data set was used for internal verification.The validation cohort was used for external verification.Results:Univariate and multivariate logistic regression analysis showed that appendiceal fecalith,periappendiceal effusion,appendiceal discontinuous pneumatosis,periappendiceal exudation were independent predictors of CA(OR=2.30,4.01,5.06,5.97).The prediction model was constructed with the abvoe 4 variables.The AUC of the model was 0.88(95%CI 0.64～0.97),and the calibration curve had a good goodness of fit(P>0.05).The AUC calculated by internal verification of the model was 0.88(95%CI 0.84～0.91),the AUC calculated by external verification was 0.88(95%CI 0.71～0.96),and the calibration curve had a good goodness of fit(P>0.05).Conclusion:The prediction model based on CT signs is valuable in diagnosing CA.
国家科技期刊平台
NSTL
万方数据
