摘要
目的:探讨基于影像组学的肺结节恶性程度预测。方法:对肺部图像影像数据库(The Lung Image Database Consortium,LIDC)-IDRI(Image Database Resource Initiative)中604例肺结节患者的CT图像进行分析,其中含肺结节的CT图像共2 803幅,医师手工勾画肺结节轮廓。根据肺结节诊断标准,共提取96个灰度、形态和纹理高通量特征,输入基于随机森林的多类分类器进行恶性程度预测。恶性程度分为5级,以数字1~5表示。随机选取1 000幅CT图像作为训练样本,剩余的1 803幅CT图像作为测试样本,实验重复10次。结果:对于单个肺结节,5类恶性程度的平均预测准确率为77.85%。对于每一类预测,曲线下面积(area under curve,AUC)均在0.94以上。对于每例患者,肺结节恶性程度的预测准确率为75.16%。结论:该研究提出的基于影像组学的方法对肺结节恶性程度的预测性能良好,可为临床诊断提供可靠的辅助信息,以利于早期发现病灶。
Objective:To investigate the method of malignancy prediction of pulmonary nodules based on radiomics.Methods:A total of2803computed tomography(CT)images containing pulmonary nodules were extracted from604scans in the publicly available dataset of The Lung Image Database Consortium(LIDC)-Image Database Resource Initiative(IDRI).Each contour of nodules was labelled by the clinical doctor.Totally96high throughput features including gray level features,shape features and texture features were extracted according to the pulmonary nodule diagnosis criteria and put into the multi-class classifier based on the random forest to predict the malignancy.The degree of malignancy was classified into1to5levels.Among all images,1000of them were randomly chosen as the training set and the rest were used as the testing set.The experiment was repeated10times.Results:For a single nodule,the average prediction accuracy of five levels was77.85%.The area under curve(AUC)of each category reached over0.94.For each patient,the malignancy prediction accuracy of pulmonary nodules was75.16%.Conclusion:The method of malignancy prediction of pulmonary nodules based on radiomics has a good performance.The results can provide a reliable basis for clinical diagnosis and help to detect the disease in the early stage.
作者
杨春然
郭翌
汪源源
YANG Chunran;GUO Yi;WANG Yuanyuan(Department of Electronic Engineering, Fudan University, Shanghai 200433, China)
出处
《肿瘤影像学》
2017年第2期97-101,共5页
Oncoradiology
