We assessed the feasibility of a data-driven imaging biomarker based on weakly supervised learning (DIB; an imaging biomarker derived from large-scale medical image data with deep learning technology) in mammography (DIB-MG). A total of 29,107 digital mammograms from five institutions (4,339 cancer cases and 24,768 normal cases) were included. After matching patients’ age, breast density, and equipment, 1,238 and 1,238 cases were chosen as validation and test sets, respectively, and the remainder were used for training. The core algorithm of DIB-MG is a deep convolutional neural network; a deep learning algorithm specialized for images. Each sample (case) is an exam composed of 4-view images (RCC, RMLO, LCC, and LMLO). For each case in a training set, the cancer probability inferred from DIB-MG is compared with the per-case ground-truth label. Then the model parameters in DIB-MG are updated based on the error between the prediction and the ground-truth. At the operating point (threshold) of 0.5, sensitivity was 75.6% and 76.1% when specificity was 90.2% and 88.5%, and AUC was 0.903 and 0.906 for the validation and test sets, respectively. This research showed the potential of DIB-MG as a screening tool for breast cancer.

Objectives The purpose of this study was to evaluate the diagnostic utility of sonographic elastography in differentiating reactive and metastatic axillary lymph nodes in breast cancer. Methods A total of 64 lymph nodes (reactive, n = 33; metastatic, n = 31) from 62 patients with breast cancer were examined by both B‐mode sonography and elastography from April to July 2009. Two experienced radiologists retrospectively assessed B‐mode sonograms by the sum of scores for 4 criteria: short diameter, shape, hilum, and cortical thickening. Elastographic images were given scores of 1 to 4 according to the percentage of high‐elasticity areas in the lymph nodes. We compared the diagnostic performance of B‐mode sonography, elastography, and combined examinations. We also calculated the strain ratio of the lymph node and subcutaneous fat tissue. Results The elasticity score for malignant lymph nodes (mean, 3.1) was higher than the score for benign lymph nodes (mean, 2.2; P < .0001). With a cutoff between elasticity scores of 2 and 3, elastography showed 80.7% sensitivity, 66.7% specificity, and 73.4% accuracy. With a cutoff between B‐mode sonographic scores of 1 and 2, B‐mode sonography showed 74.2% sensitivity and 78.8% specificity. Combined B‐mode and elastographic sonography showed higher sensitivity (87.1%) than B‐mode sonography alone. With a strain ratio cutoff point of 2.3, sensitivity was 82.8%, and specificity was 56.3%. Conclusions Sonographic elastography may increase the sensitivity of B‐mode sonography in the detection of metastatic axillary lymph nodes.

The purpose of this study was to evaluate the usefulness of applying computer-aided diagnosis (CAD) to breast ultrasound (US), depending on the operator's experience with breast imaging. Between October 2015 and January 2016, two experienced readers obtained and analyzed the grayscale US images of 200 cases according to the Breast Imaging Reporting and Data System (BI-RADS) lexicon and categories. They additionally applied CAD (S-Detect) to analyze the lesions and made a diagnostic decision subjectively, based on grayscale US with CAD. For the same cases, two inexperienced readers analyzed the grayscale US images using the BI-RADS lexicon and categories, added CAD, and came to a subjective diagnostic conclusion. We then compared the diagnostic performance depending on the operator's experience with breast imaging. The sensitivity values for the experienced readers, inexperienced readers, and CAD (for experienced and inexperienced readers) were 91.7%, 75%, 75%, and 66.7%, respectively. The specificity values for the experienced readers, inexperienced readers, and CAD (for experienced and inexperienced readers) were 76.6%, 71.8%, 78.2%, and 76.1%, respectively. When diagnoses were made subjectively in combination with CAD, the specificity significantly improved (76.6% to 80.3%) without a change in the sensitivity (91.7%) in the experienced readers. After subjective combination with CAD, the sensitivity and specificity improved in the inexperienced readers (75% to 83.3% and 71.8% to 77.1%). In addition, the area under the curve improved for both the experienced and inexperienced readers (0.84 to 0.86 and 0.73 to 0.8) after the addition of CAD. CAD is more useful for less experienced readers. Combining CAD with breast US led to improved specificity for both experienced and inexperienced readers.

Kim SH, et al. Korean J Radiol. 2013 Mar-Apr;14(2):154-163. http://dx.doi.org/10.3348/kjr.2013.14.2.154

An YY, et al. J Korean Med Sci. 2015 Jun;30(6):808-815. https://doi.org/10.3346/jkms.2015.30.6.808

This study was undertaken to compare the diagnostic performance and biopsy reduction rate of combined shear-wave elastography (SWE) and B-mode ultrasonography (US) versus B-mode US alone for breast lesions and to determine the most discriminatory parameter in SWE. A systematic review and meta-analysis were conducted. The resources for the study were obtained from MEDLINE, Embase, Cochrane Library, and KoreaMed on August 17, 2018. The quality of the articles was evaluated using the Scottish Intercollegiate Guidelines Network (SIGN) tool. Twenty-five articles with 5,147 breast lesions were selected. The meta-analysis showed pooled sensitivities of 0.94 and 0.97 (P=0.083), pooled specificities of 0.85 and 0.61 (P=0.009), and area under the receiver operating characteristic curve (AUC) of 0.96 and 0.96 (P=0.095) for combined SWE and B-mode US versus B-mode US alone. When SWE was combined with B-mode US, the Breast Imaging Reporting and Data System category changed from 4 to 3 in 71.3% of the tests, decreasing the frequency of unnecessary biopsies by 41.1%. All four parameters of SWE (the color grade of lesion stiffness, maximum elasticity, mean elasticity, and color grade of lesion stiffness/homogeneity of the lesion) improved the specificity when they were added to B-mode US. The AUC for each SWE parameter was 0.99, 0.96, 0.96, and 0.93, respectively. Adding SWE to B-mode US not only provides additional diagnostic information for differentiating between benign and malignant breast lesions, but also decreases the likelihood of unnecessary biopsies.