中国临床医学影像杂志2025，Vol.36Issue(12)：880-885,6.DOI:10.12117/jccmi.2025.12.010
双参数MRI临床-影像组学机器学习模型对移行带前列腺癌的诊断价值研究
Diagnostic value of BpMRI clinical-radiomics machine learning model for transition zone prostate cancer
摘要
Abstract
Objective:To investigate the diagnostic performance of a machine learning model that integrates biparametric magnetic resonance imaging(BpMRI)radiomics features and clinical factors for detecting prostate cancer in the transition zone(TZ-Pca).Methods:A total of 280 pathologically confirmed cases of TZ-PCa and BPH,meeting all inclusion and exclusion criteria,were retrospectively collected from two centers(center 1:203,center 2:77),and according to the proportion of 7∶3,203 patients from center 1 were completely randomly divided into training set(141 cases)and test set(62 cases).Patients from center 2 were as an independent external validation set.Independent clinical risk factors were obtained by univariate and multivariate Logistic regression analyses.Radiomics features were extracted from the largest cross-sectional area of the lesion in the T2WI and apparent diffusion coefficient(ADC)maps of the entire transition zone.The radiomics features were screened using the least absolute shrinkage and selection operator regression algorithm.Then,six machine learning algorithms were employed to construct and screen the optimal radiomics model,which was used as the basis for combining with independent clinical features to construct a clinical-radiomics combined model.The diagnostic performance of different models was compared by the area under the receiver operating characteristic(AUC)curve.The predictive efficacy and clinical value of the combined model were evaluated using calibration curve,decision curve analysis.Results:Multivariate Logistic regression analysis showed that PI-RADS scoreand tPSA were independent clinical risk factors for TZ-Pca.The SVM model based on combined BpMRI radiomics features(ADC+T2WI)performed well,with AUC of 0.865 in the training set and 0.850 in the internal validation sets,respectively.The combined model,integrating key clinical predictors,achieved AUCs of 0.963,0.889,and 0.829 in the training,internal validation,and external validation sets,respectively.Conclusions:Machine learning-based BpMRI radiomics model is useful for the diagnosis of TZ-Pca,and the combined model with clinical factors PI-RADS score and tPSA can further improve the diagnostic performance.关键词
前列腺肿瘤/磁共振成像Key words
Prostatic Neoplasms/Magnetic Resonance Imaging分类
医药卫生引用本文复制引用
YANG Jin-han,CUI Feng,JIANG Ning-ning,ZHANG Yong-sheng,XIA Lu,REN Yue,XU Hui-jing,LI Zhi-ping,WANG Jun-guang,MIAO Yu-yun..双参数MRI临床-影像组学机器学习模型对移行带前列腺癌的诊断价值研究[J].中国临床医学影像杂志,2025,36(12):880-885,6.基金项目
浙江省医药卫生科技计划项目(2025KY1161、2025KY1160、2025KY1217) （2025KY1161、2025KY1160、2025KY1217）
浙江省中医药科技计划项目(2024ZL668). （2024ZL668）
