中国医学工程2025，Vol.33Issue(5)：1-6,6.DOI:10.19338/j.issn.1672-2019.2025.05.001
MRI影像组学在直肠癌区域淋巴结良恶性鉴别中的应用研究
Application of MRI radiomics in differentiation of benign and malignant regional lymph nodes in rectal cancer
摘要
Abstract
[Objective]To investigate the application value of MRI radiomics in differentiating benign from malignant regional lymph nodes in patients with rectal cancer.By extracting and analyzing MRI image features,an effective prediction model is constructed to assist in the preoperative assessment of the metastatic status of regional lymph nodes in rectal cancer.[Methods]Data of 200 patients with rectal cancer confirmed by surgical pathology and with clear pathological diagnoses of regional lymph nodes were retrospectively analyzed.According to the status of lymph node metastasis,lymph node negativity was classified as group 1,and lymph node positivity was classified as group 2.The differences in gender and age between the two groups were compared.Feature extraction software was used to extract image feature parameters at the maximum level of lymph nodes,and the imaging omics features with obvious differences between the two groups were retained,then the optimal imaging omics features were selected to build a predictive model.All the data were divided into training set and validation set in a ratio of 7:3.Seven machine learning algorithms were used to analyze training set and validation set,the receiver operating characteristic(ROC)curve and corresponding area under the curve(AUC),specificity,sensitivity,and accuracy of the predictive model for predicting the malignancy of regional lymph nodes in rectal cancer patients were obtained,and the best prediction model based on the accuracy of machine learning models on the validation set was screened.[Results]There were 100 patients in group 1 and 100 cases in group 2.There were no statistically significant differences in gender and age between the two groups(χ2=0.09,t=1.18,P=0.772,P=0.264).A total of 286 three-dimensional texture feature parameters were extracted from the lesions,among which 87 radiomic features showed significant differences between the two groups.Ultimately,8 optimal radiomic features were retained to construct the prediction model.Support vector machine(SVM)achieved the highest accuracy in the validation set,with an AUC of 0.782 for predicting the lymph node metastasis status in the validation set.The accuracy,specificity,and sensitivity of this prediction model were 0.746,0.788,and 0.700,respectively.[Conclusion]The MRI radiomics model demonstrates its effectiveness in predicting the metastatic status of regional lymph nodes in patients with rectal cancer.关键词
直肠癌/淋巴结/磁共振成像/特征提取/机器学习Key words
rectal cancer/lymph node/magnetic resonance imaging/feature extraction/machine learning分类
临床医学引用本文复制引用
张帅,陈曦,叶钉利,黄志成..MRI影像组学在直肠癌区域淋巴结良恶性鉴别中的应用研究[J].中国医学工程,2025,33(5):1-6,6.基金项目
吉林省卫生健康科技能力提升项目(2022LC026、2022LC021) （2022LC026、2022LC021）
