AUTHOR=Zhang Renzhi , Wei Wei , Li Rang , Li Jing , Zhou Zhuhuang , Ma Menghang , Zhao Rui , Zhao Xinming 

TITLE=An MRI-Based Radiomics Model for Predicting the Benignity and Malignancy of BI-RADS 4 Breast Lesions

JOURNAL=Frontiers in Oncology

VOLUME=Volume 11 - 2021

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.733260

DOI=10.3389/fonc.2021.733260

ISSN=2234-943X

ABSTRACT=Objectives: The probability of Breast Imaging Reporting and Data Systems (BI-RADS) 4 lesions being malignant is 2%-95%, which shows the difficulty to make a diagnosis. Radiomic models based on magnetic resonance imaging (MRI) can replace clinicopathological diagnosis with high performance. In the present study, we developed and tested a radiomic model based on MRI images, that can predict the malignancy of BI-RADS 4 breast lesions.
Methods: We retrospective enrolled a total of 216 BI-RADS 4 patients MRI and clinical information. We extracted 3474 radiomics features from dynamic contrast-enhanced (DCE), T2 weighted images (T2WI), and diffusion-weighted imaging (DWI) MRI images. LASSO and logistic regression were used to select features and build radiomic models based on different sequences combinations. We built eight radiomic models which were based on DCE, DWI, T2WI, DCE+DWI, DCE+T2WI, DWI+T2WI, DCE+DWI+T2WI, and a clinical predictive model built based on the visual assessment of radiologists. A nomogram was constructed with the best radiomic signature combined with patient characteristics. The calibration curves for the radiomic signature and nomogram were conducted, combined with the Hosmer-Lemeshow test.
Results: Pearson correlation was used to eliminate 3329 irrelevant features, and then LASSO and logistic regression were used to screen the remaining feature coefficients for each model we built. Finally, 12 related features were obtained in the model which had the best performance. These 12 features were used to build a radiomic model in combination with the actual clinical diagnosis of benign or malignant lesion labels we have obtained. The best model built by 12 features from the 3 sequences has an AUC value of 0.939 (95% CI, 0.884-0.994) and an accuracy of 0.931 in the testing cohort. The sensitivity and specificity of the testing cohort are 0.932 and 0.923. The nomogram has also been verified to have calibration curves with good overlap.
Conclusions: Radiomics is beneficial in the malignancy prediction of BI-RADS 4 breast lesions. The radiomic predictive model built by the combination of DCE, DWI, and T2WI sequences has great application potential.