|本期目录/Table of Contents|

基于双序列MRI的计算机辅助诊断鉴别软组织肿瘤良恶性的研究

《现代肿瘤医学》[ISSN:1672-4992/CN:61-1415/R]

期数:
2020年12期
页码:
2122-2126
栏目:
论著（影像诊断）
出版日期:
2020-05-12

文章信息/Info

Title:
MRI-based computer-aided differentiation between benign and malignant soft-tissue tumors
作者:
姜文研1; 尚圣捷2; 王颖妮2; 王晓煜1; 宋江典3; 龙 哲2; 于 韬1; 罗娅红1
1.辽宁省肿瘤医院(中国医科大学肿瘤医院)，辽宁 沈阳 110042;2.中国医科大学生物医学工程系；3医学信息学院，辽宁 沈阳 110122
Author(s):
Jiang Wenyan1; Shang Shengjie2; Wang Yingni2; Wang Xiaoyu1; Song Jiangdian3; Long Zhe2; Yu Tao1; Luo Yahong1
1.Cancer Hospital of China Medical University,Liaoning Cancer Hospital & Institute Liaoning,Liaoning Shenyang 110042,China；2.Department of Biomedical Engineering；3.College of Medical Informatics,China Medical University,Liaoning Shenyang 110122,China.
关键词:
软组织肿瘤; MRI; 计算机辅助分析
Keywords:
soft-tissue tumors; MRI; CAD
分类号:
R738.6
DOI:
10.3969/j.issn.1672-4992.2020.12.028
文献标识码:
A
摘要:
目的：研究使用计算机辅助分析方法针对软组织肿瘤MRI影像进行肿瘤良恶性鉴别的价值。方法：回顾性收集了在辽宁省肿瘤医院就诊的72例软组织肿瘤患者的CE-T1和T1WI双序列MRI影像数据（2017年1月至2018年1月）。通过提取和筛选MRI影像特征，建立支持向量机（SVM）、K-最邻近（KNN）和随机森林（RF）三种机器学习分类器模型对肿瘤病灶进行二分类鉴别；提出一种新型集成学习分类器模型用于将两个序列MRI信息进行融合。通过绘制受试者工作特征曲线（ROC）并计算ROC曲线下面积（AUC值）以评估模型的分类鉴别能力。结果：三种机器学习分类器均取得较好的良恶性鉴别效果；提出的集成学习分类器的分类效果最佳，AUC值达到0.922（敏感性=0.965，特异性=0.783）。结论：本研究提出的计算机辅助模型能够利用MRI影像对软组织肿瘤良恶性进行有效的辅助鉴别，具有一定的潜在应用价值。
Abstract:
Objective：To analyze the identification of benign and malignant tumors using MRI images of soft tissue tumors by computer-aided diagnosis.Methods：CE-T1 and T1WI dual-sequence MRI image data were retrospectively collected from 72 soft-tissue tumors patients from Liaoning Provincial Cancer Hospital from Jan.2017 to Jan.2018.After extracting and selecting MRI image features,support vector machine(SVM),K-nearest neighbor（KNN) and random forest（RF) classifiers were established to discriminate tumor lesions.An ensemble learning classifier was proposed integrating both the CE-T1 and T1WI MRI.Operating characteristic curves（ROC) were plotted to assess the predictive abilities of the models with the area under the ROC curve（AUC).Results：The three machine learning classifiers all generated acceptable prediction results.The proposed ensemble learning classifier gave a highest AUC value of 0.922(sensitivity=0.965,specificity=0.783).Conclusion：The proposed ensemble learning classifier based on multimodal MRI data of soft-tissue tumors in this study has great potential in non-invasive distinguishing malignant from benign tumors.

参考文献/References

［1］Honore C,Faron M,Mir O,et al.Management of locoregional recurrence after radical resection of a primary nonmetastatic retroperitoneal soft tissue sarcoma：The Gustave Roussy experience［J］.J Surg Oncol,2018(118):1318-1325.
［2］LI YY,LI S.Soft tissue tumor［J］ .Journal of Jiangxi Tumor,1985(S1):36-46.［李月云,李澍.软组织肿瘤［J］.江西肿瘤,1985(S1):36-46.］
［3］Greenspan A,McGahan JP,Vogelsang P,et al.Imaging strategies in the evaluation of soft-tissue hemangiomas of the extremities：Correlation of the findings of plain radiography,angiography,CT,MRI,and ultrasonography in 12 histologically proven cases［J］.Skeletal Radiol,1992(21):11-18.
［4］Xu R,Kido S,Suga K,et al.Texture analysis on (18)F-FDG PET/CT images to differentiate malignant and benign bone and soft-tissue lesions［J］.Ann Nuc Med,2014(28)：926-935.
［5］Juntu J,Sijbers J,De Backer S,et al.Machine learning study of several classifiers trained with texture analysis features to differentiate benign from malignant soft-tissue tumors in T1-MRI images［J］.J Magn Reson Imaging,2010(31):680-689.
［6］Ezuddin NS,Pretell-Mazzini J,Yechieli RL,et al.Local recurrence of soft-tissue sarcoma：Issues in imaging surveillance strategy［J］.Skeletal Radiol,2018(47):1595-1606.
［7］Lambin P,Leijenaar RTH,Deist TM,et al.Radiomics：The bridge between medical imaging and personalized medicine［J］.Nat Rev Clin Oncol,2017(14):749-762.
［8］Wang K,Lu X,Zhou H,et al.Deep learning radiomics of shear wave elastography signifcantly improved diagnostic performance for assessing liver fbrosis in chronic hepatitis B：A prospective multicentre study［J］.Gut,2019(68):729-741.
［9］Nie K,Shi L,Chen Q,et al.Rectal cancer:Assessment of neoadjuvant chemoradiation outcome based on radiomics of multiparametric MRI［J］.Clin Cancer Res,2016(22):5256-5264.
［10］Gillies RJ,Kinahan PE,Hricak H.Radiomics:Images are more than pictures,they are data［J］.Radiology,2016,278(2):563-577.
［11］Q Xiong,X Zhou,Z Liu,et al.Multiparametric MRI-based radiomics analysis for prediction of breast cancers insensitive to neoadjuvant chemotherapy［J］.Clinical and Translational Oncology,2019(22):50-59.
［12］Wang H,Zhou Z,Li Y,et al.Comparison of machine learning methods for classifying mediastinal lymph node metastasis of non-small cell lung cancer from 18F-FDG PET/CT images［J］.EJNMMI Res,2017(7):11.
［13］Yang X,Pan X,Liu H,et al.A new approach to predict lymph node metastasis in solid lung adenocarcinoma：A radiomics nomogram［J］.Journal of Thoracic Disease,2018(10):S807-S819.
［14］Li Z,Li H,Wang S,et al.MR-Based radiomics nomogram of cervical cancer in prediction of the lymph-vascular space invasion preoperatively［J］.J Magn Reson Imaging,2019,49(5):1420-1426.
［15］Chen LD,Liang JY,Wu H,et al.Multiparametric radiomics improve prediction of lymph node metastasis of rectal cancer compared with conventional radiomics［J］.Life Sci，2018(208)：55-63.
［16］Corino VDA,Montin E,Messina A,et al.Radiomic analysis of soft tissues sarcomas can distinguish intermediate from high-grade lesions［J］.J Magn Reson Imaging，2018(47)：829-840.
［17］P Kickingereder,D Bonekamp,M Nowosielski,et al.Radiogenomics of glioblastoma：Machine learning-based classification of molecular characteristics by using multiparametric and multiregional MR imaging features［J］.Radiology,2016,281(3)：907-918.
［18］Zhu B,Chen H,Chen B,et al.Support vector machine model for diagnosing pneumoconiosis based on wavelet texture features of digital chest radiographs［J］.J Digit Imaging,2017,29(1):90-97.
［19］Farhidzadeh H,Goldgof DB,Hall LO,et al.Texture feature analysis to predict metastatic and necrotic soft tissue sarcomas［J］.SMC,2015(2015):2798-2802.
［20］Vallières M,Freeman CR,Skamene SR,et al.A radiomics model from joint FDG-PET and MRI texture features for the prediction of lung metastases in soft-tissue sarcomas of the extremities［J］.Phys Med Biol,2015(60):5471-5496.

备注/Memo

备注/Memo:
-
更新日期/Last Update: 2020-04-30