AUTHOR=Zhao Yanjie , Chen Rong , Zhang Ting , Chen Chaoyue , Muhelisa Muhetaer , Huang Jingting , Xu Yan , Ma Xuelei 

TITLE=MRI-Based Machine Learning in Differentiation Between Benign and Malignant Breast Lesions

JOURNAL=Frontiers in Oncology

VOLUME=Volume 11 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.552634

DOI=10.3389/fonc.2021.552634

ISSN=2234-943X

ABSTRACT=Background: Differential diagnosis between benign and malignant breast lesions in patients are of crucial importance for further treatments. Recent development in texture analysis and machine learning may lead to a new solution of this problem.
Method: This current study enrolled a total number of 263 patients (benign breast lesion: malignant breast lesion= 71:192) diagnosed in our hospital and received magnetic resonance imaging between January 2014 and August 2017. Patients were randomly divided into training group and validation group (4:1) and two radiologists extracted their texture features from the contrast-enhanced T1-weighted images. We performed five different feature selection methods including Histogram-based matrix (HISTO), Shape, Gray-level co-occurrence matrix (GLCM), Gray-level run length matrix (GLRLM), Gray-level zone length matrix (GLZLM), and Neighborhood gray-level dependence matrix (NGLDM) and five independent classification models were built on the basis of Linear discriminant analysis (LDA) algorithm
Results: All five models showed promising results to discriminate malignant breast lesion from benign breast lesion with the AUCs over 0.830 for both training and validation groups. The model with best discriminating ability is the combination of LDA+GBDT and the sensitivity, specificity, AUC, accuracy were 0.814, 0.779, 0.922, 0.868, respectively; LDA+RF also suggests promising results with the AUC of 0.906 in the training group.
Conclusion: The evidence from this study, while preliminary, suggests that a combination of MRI texture analysis and LDA algorithm could discriminate benign breast lesion from malignant breast lesion. Further multi-center researches in this field would be of great help in the validation of the result.