摘要
目的:讨论C T影像组学在食管鳞状细胞癌病理分化程度预测中的应用价值。方法:36例于2018年6月—2020年4月接受治疗的食管鳞状细胞癌患者随机划分成训练集与验证集,应用相关软件对图像实施特征提取及筛选处理,创建影像组学标签。结合获取的影像组学标签与临床资料展开多变量回归分析,建立影像组学模型并进行验证。通过受试者操作特征曲线评价模型术前预测食管鳞状细胞癌分化程度的效能。结果:应用影像组学构建的模型开展效能评价。在训练集中,AUC值为0.778,特异性为72.10%,灵敏度为80.37%;在验证集中,AUC值为0.755,特异性为73.38%,灵敏度为70.46%。结论:CT影像组学在食管鳞状细胞癌病理分化程度预测中的应用价值较为理想。
Objective To discuss the value of CT imaging in predicting the pathological differentiation of esophageal squamous cell carcinoma.Methods Thirty six patients with esophageal squamous cell carcinoma were randomly divided into training set and validation set.The image features were extracted and screened by relevant software,and the image omics tags were created.Multivariate regression analysis was carried out based on the obtained imaging tags and clinical data to establish and verify the imaging omics model.The performance of the model in predicting the degree of differentiation of esophageal squamous cell carcinoma(ESCC)was evaluated by receiver operating characteristic curve.Results The effectiveness evaluation was carried out by using the model constructed by imagemics.In the training set,the AUC value was 0.778,the specificity was 72.10%,and the sensitivity was 80.37%;in the validation set,the AUC value was 0.755,the specificity was 73.38%,and the sensitivity was 70.46%.Conclusion The application value of CT imaging in the prediction of pathological differentiation of esophageal squamous cell carcinoma is ideal.
作者
陈华
丁聪
张继
田为中
Chen Hua;Ding Cong;Zhang Ji;Tian Wezhong(Department of Imaging,Taizhou People's Hospital,Taizhou Jiangsu 225300,China)
出处
《影像研究与医学应用》
2020年第19期7-9,共3页
Journal of Imaging Research and Medical Applications
关键词
CT影像组学
食管鳞状细胞癌
病理分化程度
CT imaging
Esophageal squamous cell carcinoma
Degree of pathological differentiation
