Dynamic contrast-enhanced MRI intratumoral and peritumoral radiomics combined with clinical indexes for predicting Ki-67 expression of prostate cancer
10.13929/j.issn.1672-8475.2025.06.005
- VernacularTitle:动态对比增强MRI瘤内及瘤周影像组学联合临床指标预测前列腺癌Ki-67表达
- Author:
Yuanying LI
1
;
Xiaohang HU
1
;
Yajing WANG
1
;
Tong LUO
1
;
Hui LI
1
Author Information
1. 华北理工大学附属医院医学影像中心,河北 唐山 063000
- Publication Type:Journal Article
- Keywords:
prostatic neoplasms;
Ki-67 antigen;
dynamic contrast enhanced magnetic resonance imaging;
radiomics
- From:
Chinese Journal of Interventional Imaging and Therapy
2025;22(6):389-393
- CountryChina
- Language:Chinese
-
Abstract:
Objective To observe the value of intratumoral and peritumoral radiomics based on dynamic contrast-enhanced MRI(DCE-MRI)combined with clinical indexes for predicting Ki-67 expression of prostate cancer(PCa).Methods Totally 294 PCa patients were retrospectively enrolled and randomly divided into training set(n=205)and test(n=89)set at a ratio of 7∶3,who were stratified into low-expression subgroup(Ki-67≤10%)and high-expression subgroup(Ki-67＞10%)based on pathological findings.ROIintratumoral and ROIperitumoral were delineated on DCE-MRI,and radiomics features were extracted from ROIintratumora,l ROIperitumoral and ROIintratumoral+peritumora,l respectively.Radiomic features significantly associated with Ki-67 expression status were selected to construct radiomics models.A multivariate logistic regression analysis was performed to develop a clinical predicting model,then a combined model was established combined with the optimal radiomics model.Receiver operating characteristic(ROC)curves were plotted,and the area under the curve(AUC)were calculate to evaluate the predictive efficacy of the models and compared between each two models.Decision curve analysis(DCA)was used to assess the clinical net benefit of each model.Results The AUC of modelintratumora,l modelperitumoral and modelintratumoral+peritumoral for predicting Ki-67 expression in training set was 0.905,0.867 and 0.930,and the last one was the best.In clinical model,total prostate-specific antigen(t-PSA)and T stage≥3 were both independent predictors of high Ki-67 expression of PCa(both P＜0.05).The AUC of combined model based on modelintratumoral+peritumoral and clinical model was 0.911 in test set,being not significantly different with that of modelintratumoral+peritumoral(AUC=0.906;Z=0.349,P=0.727)but higher than that of clinical model(AUC=0.684;Z=4.370,P＜0.05).DCA revealed that the combined model provided higher clinical net benefit than clinical model and modelintratumoral+peritumoral across risk thresholds of 0.10-0.70.Conclusion DCE-MRI intratumoral+peritumoral radiomics model could effectively predict Ki-67 expression status of PCa.Combining with clinical indexes could further enhance its clinical net benefit.
