分子影像学杂志2025，Vol.48Issue(3)：278-283,6.DOI:10.12122/j.issn.1674-4500.2025.03.04
CT定量参数联合影像特征可预测小肠间质瘤风险分层
CT quantitative parameters combined with imaging features can predict risk classification of small intestinal stromal tumors
摘要
Abstract
Objective To explore the value of CT imaging features in predicting pathological risk classification of small intestinal stromal tumors(SIST).Methods This retrospective study analyzed 186 cases of SIST confirmed by surgical pathology at Sir Run Run Shaw Hospital from January 2016 to December 2023.The CT imaging features were analyzed,including tumor location,size,boundary,morphology,density,internal abnormal components(calcification,cystic degeneration,necrosis,surface ulceration,and intratumoral bleeding),growth pattern,enhancement pattern,feeding vessels or vessel-like enhancement,venous phase CT value,and the difference of CT values between the venous phase and plain scan.Multivariate logistic regression was employed to construct a predictive model.ROC curves were plotted,and the area under the curve(AUC)was calculated to evaluate and compare the performance of the model.Results Tumor size(OR=4.125,95%CI:2.164-7.864)and difference of CT values(OR=0.962,95%CI:0.931-0.993)were identified as independent predictors for pathological risk classification of SIST,with AUCs of 0.932(95%CI:0.891-0.973)and 0.858(95%CI:0.802-0.914),respectively.The fusion model,incorporating these two factors,demonstrated optimal predictive performance,achieving an AUC of 0.946(95%CI:0.912-0.981).Conclusion The imaging features of SIST on contrast-enhanced CT scans exhibit a certain degree of specificity.Tumor size and the difference of CT values between the venous phase and non-contrast scan serve as reliable preoperative predictors for pathological risk stratification of SIST.The fusion model,which integrates these two factors,demonstrates superior predictive performance compared to single-factor models.关键词
小肠间质瘤/胃肠道间质瘤/计算机体层成像Key words
small intestinal stromal tumor/gastrointestinal stromal tumor/computed tomography引用本文复制引用
张卜涵,彭刚,孙继红..CT定量参数联合影像特征可预测小肠间质瘤风险分层[J].分子影像学杂志,2025,48(3):278-283,6.基金项目
浙江省"领雁"研发攻关计划项目(2024C03047) （2024C03047）
