Prediction of glioma prognosis based on MR T1WI enhanced radiomics features and clinical factors in nomogram model
10.3969/j.issn.1002-1671.2025.07.005
- VernacularTitle:基于MR T1WI增强影像组学特征及临床因素的列线图模型预测胶质瘤预后
- Author:
Sheng ZHANG
1
;
Wenfeng LI
1
;
Shuo ZHUO
1
;
Jin DUAN
1
;
Jin GAO
1
;
Hong MA
1
Author Information
1. 徐州医科大学附属医院影像科,江苏 徐州 221132
- Publication Type:Journal Article
- Keywords:
glioma;
radiomics;
prognosis;
nomogram
- From:
Journal of Practical Radiology
2025;41(7):1099-1103,1113
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the value of nomogram model based on MR T1WI enhanced radiomics features and clinical fac-tors in predicting the prognosis of gliomas.Methods A retrospective selection was conducted on 135 patients with postoperative pathologically confirmed gliomas,who were categorized into poor prognosis group(n=59)and good prognosis group(n=76)according to survival condition at 20 months postoperatively.All patients were randomly divided into training group(n=94)and validation group(n=41)in a 7︰3 ratio.Radiomics features were extracted by 3D Slicer software,and the extracted radiomics features were downscaled by intraclass correlation coefficient(ICC),t-test,least absolute shrinkage and selection operator(LASSO)regression,and a total of 1 058 features were extracted for each patient,and 10 optimal radiomics features were obtained,to finally get the Radiomics score(Radscore).After combining clinical features and Radscore,a nomogram model was constructed.Results In the training group,Radscore was significantly higher in the poor prognosis group than that in the good prognosis group(t=8.773,P<0.05).The area under the curve(AUC)of receiver operating characteristic(ROC)curve of the training and validation groups of the radiomics model were 0.751[95%confidence interval(CI)0.654-0.849]and 0.606(95%CI 0.426-0.787),respectively.The AUC of the nomogram model were 0.899(95%CI 0.839-0.960)and 0.908(95%CI 0.823-0.994)in the training and validation groups,respectively,with much better predictive efficacy of the nomogram model.Conclusion A nomogram model based on MR T1WI enhanced radiomics features and clinical factors has good predictive efficacy in the prognosis of gliomas after surgery.
