Radiomics Based on Digital Mammography Helps to Identify Mammographic Masses Suspicious for Cancer

Abstract

Objectives: This study aims to build radiomics model of Breast Imaging Reporting and Data System (BI-RADS) category 4 and 5 mammographic masses extracted from digital mammography (DM) for mammographic masses characterization by using a sensitivity threshold similar to that of biopsy.

Materials and methods: This retrospective study included 288 female patients (age, 52.41 ± 10.31) who had BI-RADS category 4 or 5 mammographic masses with an indication for biopsy. The patients were divided into two temporal set (training set, 82 malignancies and 110 benign lesions; independent test set, 48 malignancies and 48 benign lesions). A total of 188 radiomics features were extracted from mammographic masses on the combination of craniocaudal (CC) position images and mediolateral oblique (MLO) position images. For the training set, Pearson's correlation and the least absolute shrinkage and selection operator (LASSO) were used to select non-redundant radiomics features and useful radiomics features, respectively, and support vector machine (SVM) was applied to construct a radiomics model. The receiver operating characteristic curve (ROC) analysis was used to evaluate the classification performance of the radiomics model and to determine a threshold value with a sensitivity higher than 98% to predict the mammographic masses malignancy. For independent test set, identical threshold value was used to validate the classification performance of the radiomics model. The stability of the radiomics model was evaluated by using a fivefold cross-validation method, and two breast radiologists assessed the diagnostic agreement of the radiomics model.

Results: In the training set, the radiomics model obtained an area under the receiver operating characteristic curve (AUC) of 0.934 [95% confidence intervals (95% CI), 0.898-0.971], a sensitivity of 98.8% (81/82), a threshold of 0.22, and a specificity of 60% (66/110). In the test set, the radiomics model obtained an AUC of 0.901 (95% CI, 0.835-0.961), a sensitivity of 95.8% (46/48), and a specificity of 66.7% (32/48). The radiomics model had relatively stable sensitivities in fivefold cross-validation (training set, 97.39% ± 3.9%; test set, 98.7% ± 4%).

Conclusion: The radiomics method based on DM may help reduce the temporarily unnecessary invasive biopsies for benign mammographic masses over-classified in BI-RADS category 4 and 5 while providing similar diagnostic performance for malignant mammographic masses as biopsies.

Keywords: BI-RADS (Breast imaging reporting and data system); Mammografy; Radiomic analysis; breast (diagnostic); breast cancer.

Figure 1

Examples of mammographic masses masking on digital mammography images. (A, C) Craniocaudal (CC) position images and mediolateral oblique position (MLO) images, respectively. (B, D) Manually drawn areas of mammographic masses masking on CC and MLO images, respectively.

Figure 2

Workflows of this study.

Figure 3

Least absolute shrinkage and selection operator (LASSO) selection process, absolute values of weights, and receiver operating characteristic curves (ROC) of 14 useful radiomics features in training set. (A) LASSO coefficient profiles of the 32 non-redundant features. The y-axis represents coefficient of each feature. The optimal value of λ was 0.0345, and the optimal log(λ) was −3.37, resulting in 14 non-zero coefficients. (B) Mean square error path using tenfold cross-validation. (C) Absolute value of weights generated by the LASSO algorithms for the optimal log(λ) value. (D) The ROC curves for a combination of shape feature, first order features, and texture features.

Figure 4

Receiver operating characteristic curves and boxplots in training and test sets. (A) Receiver operating characteristic curves (ROC) in training and test sets, with black dots representing threshold=0.22. (B) Boxplots of area under the curves (AUCs), sensitivities, and specificities generated by fivefold cross-validation in training and test sets, respectively.

Figure 5

Correlation analysis of maximum two-dimensional diameter and gray level non-uniformity (gray-level size zone matrix based) in training set. (A) Craniocaudal (CC) position images; (B) mediolateral oblique (MLO) position images. 2D, two-dimensional; GLSZM, gray-level size zone matrix.

Figure 6

Dumbbell diagram of the mean value of 14 useful radiomics features between the malignant and benign groups in the training set. SDHGLE, small-dependence high gray-level emphasis; LDLGLE, large-dependence low gray-level emphasis.