Deep learning model to predict Ki-67 expression of breast cancer using digital breast tomosynthesis.

BACKGROUND: Developing a deep learning (DL) model for digital breast tomosynthesis (DBT) images to predict Ki-67 expression. METHODS: The institutional review board approved this retrospective study and waived the requirement for informed consent from the patients. Initially, 499 patients (mean age: 50.5 years, range: 29-90 years) referred to our hospital for breast cancer were participated, 126 patients with pathologically confirmed breast cancer were selected and their Ki-67 expression measured. The Xception architecture was used in the DL model to predict Ki-67 expression levels. The high Ki-67 vs low Ki-67 expression diagnostic performance of our DL model was assessed by accuracy, sensitivity, specificity, areas under the receiver operating characteristic curve (AUC), and by using sub-datasets divided by the radiological characteristics of breast cancer. RESULTS: The average accuracy, sensitivity, specificity, and AUC were 0.912, 0.629, 0.985, and 0.883, respectively. The AUC of the four subgroups separated by radiological findings for the mass, calcification, distortion, and focal asymmetric density sub-datasets were 0.890, 0.750, 0.870, and 0.660, respectively. CONCLUSIONS: Our results suggest the potential application of our DL model to predict the expression of Ki-67 using DBT, which may be useful for preoperatively determining the treatment strategy for breast cancer.

Deep learning model for predicting the presence of stromal invasion of breast cancer on digital breast tomosynthesis.

To develop a deep learning (DL)-based algorithm to predict the presence of stromal invasion in breast cancer using digital breast tomosynthesis (DBT). Our institutional review board approved this retrospective study and waived the requirement for informed consent from the patients. Initially, 499 patients (mean age 50.5 years, age range, 29-90 years) who were referred to our hospital under the suspicion of breast cancer and who underwent DBT between March 1 and August 31, 2019, were enrolled in this study. Among the 499 patients, 140 who underwent surgery after being diagnosed with breast cancer were selected for the analysis. Based on the pathological reports, the 140 patients were classified into two groups: those with non-invasive cancer (n = 20) and those with invasive cancer (n = 120). VGG16, Resnet50, DenseNet121, and Xception architectures were used as DL models to differentiate non-invasive from invasive cancer. The diagnostic performance of the DL models was assessed based on the area under the receiver operating characteristic curve (AUC). The AUC for the four models were 0.56 [95% confidence intervals (95% CI) 0.49-0.62], 0.67 (95% CI 0.62-0.74), 0.71 (95% CI 0.65-0.75), and 0.75 (95% CI 0.69-0.81), respectively. Our proposed DL model trained on DBT images is useful for predicting the presence of stromal invasion in breast cancer.

Clinicopathological and ultrasound characteristics of breast cancer in BRCA1 and BRCA2 mutation carriers.

PURPOSE: BRCA1 and BRCA2 tumors exhibit different characteristics. This study aimed to assess and compare the ultrasound findings and pathologic features of BRCA1 and BRCA2 breast cancers. To our knowledge, this is the first study to examine the mass formation, vascularity, and elasticity in breast cancers of BRCA-positive Japanese women. METHODS: We identified patients with breast cancer harboring BRCA1 or BRCA2 mutations. After excluding patients who underwent chemotherapy or surgery before the ultrasound, we evaluated 89 cancers in BRCA1-positive and 83 in BRCA2-positive patients. The ultrasound images were reviewed by three radiologists in consensus. Imaging features, including vascularity and elasticity, were assessed. Pathological data, including tumor subtypes, were reviewed. RESULTS: Significant differences in tumor morphology, peripheral features, posterior echoes, echogenic foci, and vascularity were observed between BRCA1 and BRCA2 tumors. BRCA1 breast cancers tended to be posteriorly accentuating and hypervascular. In contrast, BRCA2 tumors were less likely to form masses. In cases where a tumor formed a mass, it tended to show posterior attenuation, indistinct margins, and echogenic foci. In pathological comparisons, BRCA1 cancers tended to be triple-negative subtypes. In contrast, BRCA2 cancers tended to be luminal or luminal-human epidermal growth factor receptor 2 subtypes. CONCLUSION: In the surveillance of BRCA mutation carriers, radiologists should be aware that the morphological differences between tumors are quite different between BRCA1 and BRCA2 patients.

The location of unilateral axillary lymphadenopathy after COVID-19 vaccination compared with that of metastasis from breast cancer without vaccination.

PURPOSE: Unilateral axillary lymphadenopathy is known to occur after coronavirus disease (COVID-19) vaccination. Post-vaccination lymphadenopathy may mimic the metastatic lymph nodes in breast cancer, and it is challenging to distinguish between them. This study investigated whether the localization of axillary lymphadenopathy on magnetic resonance imaging (MRI) could be used to distinguish reactive lymphadenopathy after COVID-19 vaccines from metastatic nodes. MATERIALS AND METHODS: We retrospectively examined preoperative MRI images of 684 axillae in 342 patients who underwent breast cancer surgery from June to October 2021. Lymphadenopathy was defined as cortical thickening or short axis ≥ 5 mm. The axilla was divided into ventral and dorsal parts on the axial plane using a perpendicular line extending from the most anterior margin of the muscle group, including the deltoid, latissimus dorsi, or teres major muscles, relative to a line along the lateral chest wall. We recorded the presence or absence of axillary lymphadenopathy in each area and the number of visible lymph nodes. RESULTS: Of 80 axillae, 41 and 39 were included in the vaccine and metastasis groups, respectively. The median time from the last vaccination to MRI was 19 days in the vaccine group. The number of visible axillary lymph nodes was significantly higher in the vaccine group (median, 15 nodes) than in the metastasis group (7 nodes) (P < 0.001). Dorsal lymphadenopathy was observed in 16 (39.0%) and two (5.1%) axillae in the vaccine and metastasis groups, respectively (P < 0.001). If the presence of both ventral and dorsal lymphadenopathy is considered indicative of vaccine-induced reaction, this finding has a sensitivity of 34.1%, specificity of 97.4%, and positive and negative predictive values of 93.3％ and 58.5%, respectively. CONCLUSION: The presence of deep axillary lymphadenopathy may be an important factor for distinguishing post-vaccination lymphadenopathy from metastasis. The number of axillary lymph nodes may also help.

Deep Learning for Differentiation of Breast Masses Detected by Screening Ultrasound Elastography.

Recently, deep learning using convolutional neural networks (CNNs) has yielded consistent results in image-pattern recognition. This study was aimed at investigating the effectiveness of deep learning using CNNs to differentiate benign and malignant breast masses identified by elastography on ultrasound screening. A data set of the elastography images of 245 breast masses (146 benign, 99 malignant) in 239 consecutive patients was retrospectively obtained. The data set was randomly split into training (55%), validation (25%) and test (20%) cohorts. A deep learning model predicting the probability of malignancy was constructed using GoogLeNet architectures (pre-trained by ImageNet) with 50 epochs. The model was then applied to the test data, and the results were compared with those obtained by evaluating the fat-to-lesion ratio (FLR) and by a 5-point visual color assessment (elasticity score). The receiver operating characteristic (ROC) curve was calculated to evaluate the performance of the model. The DeLong test was used to compare the areas under the ROC curve (AUCs). The CNN, FLR and elasticity score had a sensitivity of 0.800, 0.800 and 0.350; specificity of 0.966, 0.586 and 0.931; accuracy of 0.898, 0.673 and 0.694; positive predictive value of 0.941, 0.571 and 0.778; negative predictive value of 0.875, 0.810 and 0.675; and AUC of 0.895, 0.693 and 0.641, respectively. The AUC of the CNN was significantly higher than that of the FLR or elasticity score (p < 0.001). A CNN-based deep learning model for predicting benign or malignant breast masses revealed better diagnostic performance than did FLR or elasticity score-based estimations on ultrasound elastography. The CNN-based model also increased the positive predictive value from 57%-78% to 94%. Therefore, this model may reduce unnecessary biopsy recommendations for masses detected on breast ultrasound screening.

Why is the depth/width ratio of a typical fibroadenoma small in breast ultrasonography?

PURPOSE: Typical myxomatous fibroadenomas have a small depth/width (D/W) ratio on ultrasonography. The small D/W ratio of fibroadenomas is speculated to be caused by the softness of the mass and its orientation along the longitudinal aspect of the ductal elements without adhesion to the surrounding tissue; however, this has not been clearly proven. This study aimed to confirm the reason why fibroadenomas present with a small D/W ratio on ultrasonography. METHODS: We retrospectively analyzed imaging data from 17 patients who were diagnosed with typical fibroadenomas on ultrasonography and who underwent magnetic resonance imaging (MRI) at our hospital. RESULTS: The median D/W ratio obtained from ultrasonography images was 0.48 (0.32-0.67), while that obtained from MRI was 1.38 (0.62-1.68). The D/W ratios calculated from MRI were significantly greater than those calculated from ultrasonography images (p < 0.001). The D/W ratio obtained using ultrasonography was not greater than the D/W ratio obtained using MRI in any of the cases. CONCLUSION: This study revealed that the small D/W ratio of fibroadenomas on ultrasonography may be attributable to the horizontal force acting on the breast against the chest wall in the supine position, the elasticity of the fibroadenoma, and the lack of adhesion between the mass and surrounding tissue.

Deep learning model for breast cancer diagnosis based on bilateral asymmetrical detection (BilAD) in digital breast tomosynthesis images.

The purpose of this study was to develop a deep learning model to diagnose breast cancer by embedding a diagnostic algorithm that examines the asymmetry of bilateral breast tissue. This retrospective study was approved by the institutional review board. A total of 115 patients who underwent breast surgery and had pathologically confirmed breast cancer were enrolled in this study. Two image pairs [230 pairs of bilateral breast digital breast tomosynthesis (DBT) images with 115 malignant tumors and contralateral tissue (M/N), and 115 bilateral normal areas (N/N)] were generated from each patient enrolled in this study. The proposed deep learning model is called bilateral asymmetrical detection (BilAD), which is a modified convolutional neural network (CNN) model of Xception with two-dimensional tensors for bilateral breast images. BilAD was trained to classify the differences between pairs of M/N and N/N datasets. The results of the BilAD model were compared to those of the unilateral control CNN model (uCNN). The results of BilAD and the uCNN were as follows: accuracy, 0.84 and 0.75; sensitivity, 0.73 and 0.58; and specificity, 0.93 and 0.92, respectively. The mean area under the receiver operating characteristic curve of BilAD was significantly higher than that of the uCNN (p = 0.02): 0.90 and 0.84, respectively. The proposed deep learning model trained by embedding a diagnostic algorithm to examine the asymmetry of bilateral breast tissue improves the diagnostic accuracy for breast cancer.

Relationship Between Breast Density, Breast Cancer Subtypes, and Prognosis.

BACKGROUND: In the United States, Europe, and Asia, a consensus has been reached that there is a higher risk of breast cancer in high density breasts. However, there are some contrary reports that suggest the absence of an association between breast composition and breast cancer subtype; thus, there is conflicting evidence. The purpose of this study was to investigate trends in the incidence of breast cancer subtypes according to breast composition and analyze the survival rates in Japanese women. PATIENTS AND METHODS: Between 2007 and 2008, 1258 Japanese patients with invasive breast cancer who underwent mammography and obtained a pathological diagnosis in our institution were included in the study. We compared cancer subtypes with breast composition types (dense and non-dense breast), and classified them based on initial mammography findings. Information on 5- and 10-year survival rates was collected by chart review for patients with dense and nondense breasts. Statistical analysis was performed using the Pearson's chi-square test for breast composition and cancer subtype. The effect of breast composition on mortality was examined using a multivariate Cox proportional hazards model, and adjusted hazard ratios were calculated. RESULTS: No significant difference was found between breast cancer subtype and breast composition (P = .08). Five-year (log-rank test, P = .09) and 10-year (log-rank test, P = .31) survival rates were not significantly different between breast composition types. CONCLUSION: There was no significant association between breast composition and cancer subtypes. There was also no significant difference in the prognosis between patients with and without dense breasts.