A Convolutional Neural Network Based on Ultrasound Images of Primary Breast Masses: Prediction of Lymph-Node Metastasis in Collaboration With Classification of Benign and Malignant Tumors.

Purpose: A convolutional neural network (CNN) can perform well in either of two independent tasks [classification and axillary lymph-node metastasis (ALNM) prediction] based on breast ultrasound (US) images. This study is aimed to investigate the feasibility of performing the two tasks simultaneously. Methods: We developed a multi-task CNN model based on a self-built dataset containing 5911 breast US images from 2131 patients. A hierarchical loss (HL) function was designed to relate the two tasks. Sensitivity, specificity, accuracy, precision, F1-score, and analyses of receiver operating characteristic (ROC) curves and heatmaps were calculated. A radiomics model was built by the PyRadiomics package. Results: The sensitivity, specificity and area under the ROC curve (AUC) of our CNN model for classification and ALNM tasks were 83.5%, 71.6%, 0.878 and 76.9%, 78.3%, 0.836, respectively. The inconsistency error of ALNM prediction corrected by HL function decreased from 7.5% to 4.2%. Predictive ability of the CNN model for ALNM burden (≥3 or ≥4) was 77.3%, 62.7%, and 0.752, and 66.6%, 76.8%, and 0.768, respectively, for sensitivity, specificity and AUC. Conclusion: The proposed multi-task CNN model highlights its novelty in simultaneously distinguishing breast lesions and indicating nodal burden through US, which is valuable for "personalized" treatment.

Value of conventional ultrasound, ultrasound elasticity imaging, and acoustic radiation force impulse elastography for prediction of malignancy in breast lesions.

OBJECTIVE: To evaluate the efficacies of conventional ultrasound (US), US elasticity imaging (EI), and acoustic radiation force impulse (ARFI) elastography in breast malignancy diagnosis. METHODS: We included 315 women (mean age, 44 years; range, 18-81 years) with 336 pathologically proven breast lesions in this retrospective study. All lesions underwent conventional US, EI, and ARFI (including virtual touch tissue imaging [VTI], virtual touch tissue quantification [VTQ], and virtual touch tissue imaging and quantification [VTIQ]) elastography. Multivariate logistic regression analysis was performed to assess 12 independent variables for malignancy prediction. Diagnostic performance was evaluated with receiver operating characteristic (ROC) curve analysis. RESULTS: Irregular lesion shape was the strongest independent predictor for breast malignancy, followed by poorly defined margins, taller than wide dimensions, posterior echo attention, VTIQ, and VTI boundaries (P < 0.05). Area under the ROC curve (AUC) for VTIQ was higher than other significant independent variables. With the best cut-off value of 3.74 m/s, the AUC, sensitivity, and specificity were 0.93 (95% CI: 0.90, 0.96), 90.1%, and 91.1%, respectively. CONCLUSIONS: ARFI elastography is a promising method in breast malignancy prediction, with good diagnostic performance. For patients requiring surgery, the combination of various methods can provide better diagnostic results and may help to reduce unnecessary biopsy or surgery.