Prediction of Ki-67 expression level in breast cancer based on double phase DCE-MRI radiomics combined with clinico-radiological characteristics
ZHANG Saisai
RONG Jing
SHAO Min
YIN Likang
WANG Min
XU Yongsheng
WANG Xiao
Abstract：Objective To investigate the clinical value of combined model based on clinical and double phase dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI)radiomics features in predicting the expression level of Ki-67 in breast cancer.Methods A total of 155 patients with confirmed breast cancer at the First Affiliated Hospital of Anhui Medical University from January 2021 to August 2023 were retrospectively collected and categorized into high expression group(Ki-67≥30%,n=84)and low expression group(Ki-67<30%,n=71)according to postoperative immunohistoche-misty results.All cases were then randomly assigned to either a training set(n=108)or a test set(n=47)at a ratio of 7:3.The clinical model,the radiomics model(enhanced early and later phase),and the combined model were constructed using the selected clinico-radiological and radiomics features(enhanced early and later phase).ROC curves were used to evaluate the diagnostic efficacy of the four models.Calibration curves and decision curves were subsequently employed to assess the clinical utility of the predictive models.Results Ultimately,three clinico-radiological features and seven radiomics features were selected.The model constructed by combining the radiomics features with the clinico-radiological features showed various improvements in the effectiveness of Ki-67 expression prediction.The AUC of the combined model were 0.924 and 0.909 in the training and test sets,respectively.Calibration and decision curves showed that the combined model had promising clinical application.Conclusion The model based on double phase DCE-MRI radiomics features combined with clinico-radiological features has high predictive efficacy for Ki-67 expression in breast cancer.
Keywords：radiomicsbreast cancerKi-67dynamic contrast-enhanced magnetic resonance imagingpreoperative prediction
Publication Date：2025-07-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：9( 855-863 )