Breast Imaging Reporting Research Articles

Development and internal validation of a conventional ultrasound-based nomogram for predicting malignant nonmasslike breast lesions

The aim of this study was to develop a conventional ultrasound (US) features-based nomogram for the prediction of malignant nonmasslike (NML) breast lesions. Consecutive cases of adult females diagnosed with NML breast lesions via US screening in our center from June 1st, 2017, to April 17th, 2020, were retrospectively enrolled. Candidate variables included age, clinical symptoms, and the image features obtained from the conventional US. Nomograms were developed based on the results of the multiple logistic regression analysis via R language. One thousand bootstraps were used for internal validation. The area under the curve (AUC) and the bias-corrected concordance index (C-index) were calculated. Decision curve analysis (DCA) was also performed for further comparison between the nomogram and the Breast Imaging Reporting and Data System (BI-RADS). The study has not yet been registered. A total of 229 patients were included in the study after exclusion and follow-up. The overall malignant rate of NML breast lesions was 31.0%. Age, clinical symptoms, echo pattern, calcification, orientation, and Adler's classification were selected to generate the nomogram according to the results of the multivariable logistic regression analysis. The bias-corrected C-index and the AUC of our nomogram were 0.790 and 0.828, respectively. The DCA showed that our model had larger net benefits in a range from 0.2 to 0.7 when compared with the BI-RADS. We developed a prediction model using a combination of age, clinical symptoms, echo pattern, calcification, orientation, and Adler's classification for malignant NML breast lesion prediction that yielded adequate discrimination and calibration.

Mammographic breast density status in women aged more than 40 years in Sulaimaniyah, Iraq: a cross-sectional study.

Mammography is the gold standard screening procedure for the early diagnosis of breast cancer. This study aimed to determine the distribution of breast density among women older than 40 years in Sulaimaniyah, Iraq, and to examine the correlations between breast density and various risk factors. This cross-sectional study included 750 women who received routine mammographic breast screening at Sulaimaniyah Breast Center. Bilateral standard two-view mammographic images (craniocaudal and mediolateral oblique projections) were acquired and reported using a picture archiving and communication system. American College of Radiology (ACR) Breast Imaging-Reporting and Data System (BI-RADS) assessment categories C and D were considered as dense. A total of 54.3% of breasts were classified as dense, with ACR-BI-RADS categories C or D. Breast density was significantly associated with age, body mass index, a family history of breast cancer, and pre-menopause, and women with no history of breastfeeding were more likely to have dense breasts than those with partial or complete breastfeeding. This study revealed that women from Sulaimaniyah with a distinct breast-density profile at mammographic screening may have a significantly increased risk of breast cancer.

Contrast Enhanced Mammography Descriptors using Breast Imaging and Reporting and Data System

AbstractContrast-enhanced mammography (CEM) is an emerging breast imaging technique with a potential to develop into a one-stop imaging solution for several conditions which require imaging, especially its role in cancer detection and local staging when combined with ultrasound. In 2022, CEM lexicon was published as a supplement to the American College of Radiology Breast Imaging Reporting and Data System (ACR BIRADS) fifth edition. This pictorial review illustrates the findings that are seen on CEM using ACR BIRADS CEM lexicon to make the reader familiar with the terminologies. In our experience, interpretation and description of findings in CEM using ACR-BIRADS CEM lexicon had a short learning curve and it was easy to implement in practice as most of the terminologies were already in use for interpreting magnetic resonance imaging.

Upgrade Rate and Predictive Factors Associated With Breast Papillary Lesions on Core Biopsy: A Canadian Experience.

Background: Papillary lesions of the breast are a heterogeneous group, encompassing a wide range of lesions. The histologic distinction between papillary breast lesions remains challenging, especially on core biopsy specimens. Aim: This study aimed to determine the rate of upgrade to atypia or malignancy of biopsy-proven papillary lesions on surgical follow-up and to assess for factors associated with an upgrade in Greater Vancouver, BC, Canada. Materials and Methods: This is a retrospective population-based study of all breast papillary lesions diagnosed on core biopsy between 2017 and 2019 in the Fraser Health Authority in Greater Vancouver, Canada. Patients were retrieved from the laboratory information system. Patient demographics, histopathologic, and radiologic findings were analyzed. Results: A total of 269 specimens from 269 patients (mean 61.1 years), including 265 female and 4 male patients, were included in the study. Of the 269 specimens, 129 (48%) were intraductal papillomas and 140 (52%) were atypical papillary lesions. The overall upgrade rate among papillomas was 11.6% (15 of 129) on final excision. The mean age of patients diagnosed with papilloma on core biopsy was significantly younger than those with atypical papillary lesions (55.6 vs 66.1 years, P < .0001). Lesion size in patients with papillomas on core biopsy was significantly smaller than those with atypical papillary lesions (11.1 vs 15.1 mm, P = .001). The upgrade rates in patients <55 and ≥55 years were 4.9% and 13.2%. Size (P = .004) and atypia on core biopsy (P = .009) were significantly associated with upgrade. Older age (>55 years) (OR = 5.3, 95% CI: 1.04-27.08) was an independent predictor of upgrade among papillomas. Size, location, and Breast Imaging-Reporting and Data System (BI-RADS) radiologic categories in our study were not associated with predicting the upgrade of papillomas. Conclusion: Our data suggest that the risk of upgrade to atypia or malignancy is sufficient to warrant the excision of benign papillomas of any size in patients aged ≥55 years. In patients younger than 55 years, observation with close clinical and radiological follow-up without surgery may be sufficient. Our findings also support surgical excision of papillomas diagnosed on core biopsy when associated with atypia.

Reproducibility assessment of uptake on dedicated breast PET for noise discrimination.

Dedicated breast PET (dbPET) systems have improved the detection of small breast cancers but have increased false-positive diagnoses due to an increased chance of noise detection. This study examined whether reproducibility assessment using paired images helped to improve noise discrimination and diagnostic performance in dbPET. This study included 21 patients with newly diagnosed breast cancer who underwent [18F]FDG-dbPET and contrast-enhanced breast MRI. A 10-min dbPET data scan was acquired per breast, and two sets of reconstructed images were generated (named dbPET-1 and dbPET-2, respectively), each of which consisted of randomly allocated 5-min data from the 10-min data. Uptake spots higher than the background were indexed for the study with visual assessment. All indexed uptakes on dbPET-1 were evaluated using dbPET-2 for reproducibility. MRI findings based on the Breast Imaging-Reporting and Data System (BI-RADS) 2013 were used as the gold standard. Uptake spots that corresponded to BI-RADS 1 on MRI were considered noise, while those with BI-RADS 4b-6 were considered malignancies. The diagnostic performance of dbPET for malignancy was evaluated using four different criteria: any uptake on dbPET-1 regarded as positive (criterion A), a subjective visual assessment of dbPET-1 (criterion B), reproducibility assessment between dbPET-1 and dbPET-2 (criterion C), and a combination of B and C (criterion D). A total of 213 indexed uptake spots were identified on dbPET-1, including 152, 15, 6, 6, and 34 lesions classified as BI-RADS MRI categories 1, 2, 4b, 4c, and 5, respectively. Overall, 31.9% of the index uptake values were reproducible. All malignant lesions were reproducible, whereas 93.4% of noise was not reproducible. The sensitivities for malignancy for criteria A, B, C, and D were 100%, 91.3%, 100%, and 91.3%, respectively, with positive predictive values (PPVs) of 21.4%, 68.9%, 67.6%, and 82.4%, respectively. Our results demonstrated that reproducibility assessment helped reduce false-positive findings caused by noise on dbPET without lowering the sensitivity for malignancy. While subjective visual assessment was also efficient in increasing PPV, it occasionally missed malignant uptake.

Evaluation of diagnostic efficacy of multimode ultrasound in BI-RADS 4 breast neoplasms and establishment of a predictive model.

To explore the diagnostic efficacy of ultrasound (US), two-dimensional and three-dimensional shear-wave elastography (2D-SWE and 3D-SWE), and contrast-enhanced ultrasound (CEUS) in breast neoplasms in category 4 based on the Breast Imaging Reporting and Data System (BI-RADS) from the American College of Radiology (ACR) and to develop a risk-prediction nomogram based on the optimal combination to provide a reference for the clinical management of BI-RADS 4 breast neoplasms. From September 2021 to April 2022, a total of 104 breast neoplasms categorized as BI-RADS 4 by US were included in this prospective study. There were 78 breast neoplasms randomly assigned to the training cohort; the area under the receiver-operating characteristic curve (AUC), 95% confidence interval (95% CI), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 2D-SWE, 3D-SWE, CEUS, and their combination were analyzed and compared. The optimal combination was selected to develop a risk-prediction nomogram. The performance of the nomogram was assessed by a validation cohort of 26 neoplasms. Of the 78 neoplasms in the training cohort, 16 were malignant and 62 were benign. Among the 26 neoplasms in the validation cohort, 6 were malignant and 20 were benign. The AUC values of 2D-SWE, 3D-SWE, and CEUS were not significantly different. After a comparison of the different combinations, 2D-SWE+CEUS showed the optimal performance. Least absolute shrinkage and selection operator (LASSO) regression was used to filter the variables in this combination, and the variables included Emax, Eratio, enhancement mode, perfusion defect, and area ratio. Then, a risk-prediction nomogram with BI-RADS was built. The performance of the nomogram was better than that of the radiologists in the training cohort (AUC: 0.974 vs. 0.863). In the validation cohort, there was no significant difference in diagnostic accuracy between the nomogram and the experienced radiologists (AUC: 0.946 vs. 0.842). US, 2D-SWE, 3D-SWE, CEUS, and their combination could improve the diagnostic efficiency of BI-RADS 4 breast neoplasms. The diagnostic efficacy of US+3D-SWE was not better than US+2D-SWE. US+2D-SWE+CEUS showed the optimal diagnostic performance. The nomogram based on US+2D-SWE+CEUS performs well.

Clinical use and adjustment of ultrasound elastography for breast lesions followed WFUMB guidelines and recommendations in the real world.

This study aimed to explore the value of strain elastography (SE) and shear wave elastography (SWE) following the World Federation of Ultrasound in Medicine and Biology (WFUMB) guidelines and recommendations in the real world in distinguishing benign and malignant breast lesions and reducing biopsy of BI-RADS (Breast Imaging Reporting and Data System) 4a lesions. This prospective study included 274 breast lesions. The elastography score (ES) by the Tsukuba score, the strain ratio (SR) for SE, and Emax for SWE of the lesion(A) and the regions(A') included the lesion and the margin (0.5-5mm) surrounding the lesion were measured. The sensitivity, specificity, and AUC were calculated and compared by the cutoff values recommended by WFUMB guidelines. When scores of 1 to 3 were classified as probably benign by WFUMB recommendation, the ES was significantly higher in malignant lesions compared to benign lesions (p < 0.05) in all lesions. For the cohort by size >20mm, the sensitivity was 100%, and the specificity was 45.5%. ES had the highest AUC: 0.79(95% CI 0.72-0.86) with a sensitivity of 96.2%, and a specificity of 61.8% for the cohort by size ≤20 mm. For the Emax-A'-S2.5mm, when the high stiffness would be considered with Emax above 80 kPa in SWE, the malignant lesions were diagnosed with a sensitivity of 95.8%, a specificity of 43.3% for all lesions, a sensitivity of 88.5% for lesions with size ≤20 mm, and sensitivity of 100.0% for lesions with size >20mm. In 84 lesions of BI-RADS category 4a, if category 4a lesions with ES of 1-3 points or Emax-A'-S2.5 less than 80 kPa could be downgraded to category 3, 52 (61.9%) lesions could be no biopsy, including two malignancies. If category 4a lesions with ES of 1-3 points and Emax-A'-S2.5 less than 80kPa could be downgraded to category 3, 23 (27.4%) lesions could be no biopsy, with no malignancy. The elastography score for SE and Emax-A' for SWE after our modification were beneficial in the diagnosis of breast cancer. The combination of SWE and SE could effectively reduce the biopsy rate of BI-RADS category 4a lesions.

Comparison of diagnostic potential of shear wave elastography between breast mass lesions and non–mass–like lesions

BackgroundShear wave elastography (SWE) can improve the specificity of B-mode ultrasound (US) without reducing the sensitivity for breast cancer diagnosis. Existing research on SWE includes both mass lesions and non‑mass‑like (NML) lesions or only NML lesions; however, there are no studies comparing the diagnostic potential of SWE in the detection of mass and NML lesions in the same trial. ObjectiveThis study aimed to compare the diagnostic performance of SWE in detecting mass lesions and NML lesions and determine the different individualised thresholds of the SWE parameters according to the lesion type. MethodsThis Study included 623 breast lesions of 562 consecutive women, who were scheduled for conventional US and SWE between January 2021 and December 2021. The diagnostic performances of conventional US and each quantitative SWE parameter (maximum elastic modulus [Emax], mean elastic modulus [Emean], and elastic modulus standard deviation [Esd]) were assessed. Histological diagnosis for all Breast Imaging Reporting and Database System (BI-RADS) category 4/5 patients and some BI-RADS category 3 patients and the follow-up results of other BI-RADS category 3 patients were used as the reference standard. ResultsIn this study, 281 benign lesions and 342 malignant lesions were identified. The diagnostic performance of conventional US and SWE was better in the mass lesion group than in the NML lesion group. Every SWE parameter had a different threshold in each group, and the thresholds of the SWE parameters were higher in the mass lesion group than in the NML lesion group. In the mass lesion group, Esd had the highest Az value, whereas in the NML lesion group, Emax had the highest Az value. In both the mass and NML lesion groups, the diagnostic specificity of the combination of conventional US and SWE was significantly higher than that of conventional US alone (P < 0.05), without a significantly decrease in the diagnosticsensitivity. ConclusionsSWE could increase the confidence of breast ultrasound diagnosis, especially for NML lesions. NML lesions had lower thresholds of SWE parameters than did the mass lesions.

Precision Medicine in Breast Cancer: Do MRI Biomarkers Identify Patients Who Truly Benefit from the Oncotype DX Recurrence Score® Test?

The aim of this study was to combine breast MRI-derived biomarkers with clinical-pathological parameters to identify patients who truly need an Oncotype DX Breast Recurrence Score® (ODXRS) genomic assay, currently used to predict the benefit of adjuvant chemotherapy in ER-positive/HER2-negative early breast cancer, with the ultimate goal of customizing therapeutic decisions while reducing healthcare costs. Patients who underwent a preoperative multiparametric MRI of the breast and ODXRS tumor profiling were retrospectively included in this study. Imaging sets were evaluated independently by two breast radiologists and classified according to the 2013 American College of Radiology Breast Imaging Reporting and Data System (ACR BI-RADS) lexicon. In a second step of the study, a combined oncologic and radiologic assessment based on clinical-pathological and radiological data was performed, in order to identify patients who may need adjuvant chemotherapy. Results were correlated with risk levels expressed by ODXRS, using the decision made on the basis of the ODXRS test as a gold standard. The χ2 test was used to evaluate associations between categorical variables, and significant ones were further investigated using logistic regression analyses. A total of 58 luminal-like, early-stage breast cancers were included. A positive correlation was found between ODXRS and tumor size (p = 0.003), staging (p = 0.001) and grading (p = 0.005), and between BI-RADS categories and ODXRS (p &lt; 0.05 for both readers), the latter being confirmed at multivariate regression analysis. Moreover, BI-RADS categories proved to be positive predictors of the therapeutic decision taken after performing an ODXRS assay. A statistically significant association was also found between the therapeutic decision based on the ODXRS and the results of combined onco-radiologic assessment (p &lt; 0.001). Our study suggests that there is a correlation between BI-RADS categories at MRI and ODXRS and that a combined onco-radiological assessment may predict the decision made on the basis of the results of ODXRS genomic test.

Improving breast cancer diagnosis by incorporating raw ultrasound parameters into machine learning

The improved diagnostic accuracy of ultrasound breast examinations remains an important goal. In this study, we propose a biophysical feature-based machine learning method for breast cancer detection to improve the performance beyond a benchmark deep learning algorithm and to furthermore provide a color overlay visual map of the probability of malignancy within a lesion. This overall framework is termed disease-specific imaging. Previously, 150 breast lesions were segmented and classified utilizing a modified fully convolutional network and a modified GoogLeNet, respectively. In this study multiparametric analysis was performed within the contoured lesions. Features were extracted from ultrasound radiofrequency, envelope, and log-compressed data based on biophysical and morphological models. The support vector machine with a Gaussian kernel constructed a nonlinear hyperplane, and we calculated the distance between the hyperplane and each feature’s data point in multiparametric space. The distance can quantitatively assess a lesion and suggest the probability of malignancy that is color-coded and overlaid onto B-mode images. Training and evaluation were performed on in vivo patient data. The overall accuracy for the most common types and sizes of breast lesions in our study exceeded 98.0% for classification and 0.98 for an area under the receiver operating characteristic curve, which is more precise than the performance of radiologists and a deep learning system. Further, the correlation between the probability and Breast Imaging Reporting and Data System enables a quantitative guideline to predict breast cancer. Therefore, we anticipate that the proposed framework can help radiologists achieve more accurate and convenient breast cancer classification and detection.

Application of Deep Learning to Reduce the Rate of Malignancy Among BI-RADS 4A Breast Lesions Based on Ultrasonography.

The aim of the work described here was to develop an ultrasound (US) image-based deep learning model to reduce the rate of malignancy among breast lesions diagnosed as category 4A of the Breast Imaging-Reporting and Data System (BI-RADS) during the pre-operative US examination. A total of 479 breast lesions diagnosed as BI-RADS 4A in pre-operative US examination were enrolled. There were 362 benign lesions and 117 malignant lesions confirmed by postoperative pathology with a malignancy rate of 24.4%. US images were collected from the database server. They were then randomly divided into training and testing cohorts at a ratio of 4:1. To correctly classify malignant and benign tumors diagnosed as BI-RADS 4A in US, four deep learning models, including MobileNet, DenseNet121, Xception and Inception V3, were developed. The performance of deep learning models was compared using the area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Meanwhile, the robustness of the models was evaluated by five-fold cross-validation. Among the four models, the MobileNet model turned to be the optimal model with the best performance in classifying benign and malignant lesions among BI-RADS 4A breast lesions. The AUROC, accuracy, sensitivity, specificity, PPV and NPV of the optimal model in the testing cohort were 0.897, 0.913, 0.926, 0.899, 0.958 and 0.784, respectively. About 14.4% of patients were expected to be upgraded to BI-RADS 4B in US with the assistance of the MobileNet model. The deep learning model MobileNet can help to reduce the rate of malignancy among BI-RADS 4A breast lesions in pre-operative US examinations, which is valuable to clinicians in tailoring treatment for suspicious breast lesions identified on US.

Automated Registration for Dual-View X-Ray Mammography Using Convolutional Neural Networks.

Automated registration algorithms for a pair of 2D X-ray mammographic images taken from two standard imaging angles, namely, the craniocaudal (CC) and the mediolateral oblique (MLO) views, are developed. A fully convolutional neural network, a type of convolutional neural network (CNN), is employed to generate a pixel-level deformation field, which provides a mapping between masses in the two views. Novel distance-based regularization is employed, which contributes significantly to the performance. The developed techniques are tested using real 2D mammographic images, slices from real 3D mammographic images, and synthetic mammographic images. Architectural variations of the neural network are investigated and the performance is characterized from various aspects including image resolution, breast density, lesion size, lesion subtlety, and lesion Breast Imaging-Reporting and Data System (BI-RADS) category. Our network outperformed the state-of-the-art CNN-based and non-CNN-based registration techniques, and showed robust performance across various tissue/lesion characteristics. The proposed methods provide a useful automated tool for co-locating lesions between the CC and MLO views even in challenging cases. Our methods can aid clinicians to establish lesion correspondence quickly and accurately in the dual-view X-ray mammography, improving diagnostic capability.

Association of the International Academy of Cytology category with the Breast Imaging Reporting and Data System score in relation to the diagnostic accuracy for breast lumps

Background: Breast lumps are the commonest patient presentation in breast clinics. Initial diagnostic approaches, such as mammography and fine needle aspiration cytology (FNAC), help in prompting proper preoperative diagnosis and planning management. Methods: This retrospective study included all patients who presented with breast lumps and underwent breast imaging (with Breast Imaging Reporting and Data System [BIRADS] scoring) and FNAC (with International Academy of Cytology [IAC] category assessment) in our hospital. The IAC category was compared with the BIRADS score. The FNAC and mammographic findings were compared with available histopathological findings. Data were entered into MS Excel worksheets, and statistical analysis was conducted using SPSS 22.0. Results: A total of 60 patients were included in the study. The mean patient age was 40.11 years. Of the 60 cases, 35 benign cases and 15 malignant cases were concordant with the diagnosis according to the BIRADS score and IAC category. The association between the BIRADS score and IAC category was significant (p = 0.004). The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the BIRADS score in diagnosing breast lesions were 85.37%, 78.95%, 89.74%, 71.43%, and 83.33%, respectively, compared with those of the IAC category. The sensitivity, specificity, PPV, NPV, and accuracy of the IAC category were 82.35%, 92.31%, 93.33%, 80.0%, and 86.67%, respectively, compared with those of the available histopathological findings. The sensitivity, specificity, PPV, NPV, and accuracy of the BIRADS score were 76.92%, 76.47%, 71.43%, 81.25%, and 76.67%, respectively, compared with those of the available histopathological findings. Conclusion: FNAC with IAC category assessment and mammography with BIRADS scoring can be used as first-line diagnostic tests for breast lumps. However, FNAC is more sensitive in diagnosing breast lumps than mammography.

Algorithmic transparency and interpretability measures improve radiologists’ performance in BI-RADS 4 classification

ObjectiveTo evaluate the perception of different types of AI-based assistance and the interaction of radiologists with the algorithm’s predictions and certainty measures.MethodsIn this retrospective observer study, four radiologists were asked to classify Breast Imaging-Reporting and Data System 4 (BI-RADS4) lesions (n = 101 benign, n = 99 malignant). The effect of different types of AI-based assistance (occlusion-based interpretability map, classification, and certainty) on the radiologists’ performance (sensitivity, specificity, questionnaire) were measured. The influence of the Big Five personality traits was analyzed using the Pearson correlation.ResultsDiagnostic accuracy was significantly improved by AI-based assistance (an increase of 2.8% ± 2.3%, 95 %-CI 1.5 to 4.0 %, p = 0.045) and trust in the algorithm was generated primarily by the certainty of the prediction (100% of participants). Different human-AI interactions were observed ranging from nearly no interaction to humanization of the algorithm. High scores in neuroticism were correlated with higher persuasibility (Pearson’s r = 0.98, p = 0.02), while higher consciousness and change of accuracy showed an inverse correlation (Pearson’s r = −0.96, p = 0.04).ConclusionTrust in the algorithm’s performance was mostly dependent on the certainty of the predictions in combination with a plausible heatmap. Human-AI interaction varied widely and was influenced by personality traits.Key Points• AI-based assistance significantly improved the diagnostic accuracy of radiologists in classifying BI-RADS 4 mammography lesions.• Trust in the algorithm’s performance was mostly dependent on the certainty of the prediction in combination with a reasonable heatmap.• Personality traits seem to influence human-AI collaboration. Radiologists with specific personality traits were more likely to change their classification according to the algorithm’s prediction than others.

Value of multimodal imaging in the diagnosis of breast sclerosing adenosis associated with malignant lesions.

To explore the diagnostic value of multimodal imaging techniques, including automatic breast volume scanner (ABVS), mammography (MG), and magnetic resonance (MRI) in breast sclerosing adenosis (SA) associated with malignant lesions. From January 2018 to October 2020, 76 patients (88 lesions) with pathologically confirmed as SA associated with malignant or benign lesions were retrospective analyzed. All patients completed ABVS examination, 58 patients (67 lesions) with MG and 50 patients (62 lesions) with MRI were also completed before biopsy or surgical excision, of which, six patients (eight lesions) diagnosed as Breast Imaging Reporting and Data System (BI-RADS) category 3 by all imaging examinations underwent surgical excision without biopsy, other 70 patients (80 lesions) with BI-RADS category 4 or above by any imaging examination completed biopsy, including 65 patients (75 lesions) were further surgical excised and the other five patients (five lesions) were just followed up. All lesions were retrospectively described and classified, and were divided into benign group and malignant group according to their pathological results. Image features of different examination methods between the two groups were compared and analyzed. A ROC curve was established using the sensitivity of BI-RADS categories to predict malignant lesions in different imaging techniques as the ordinate and 1-specificity as the abscissa. 88 lesions including 26 purely SA and 45 SA associated with benign lesions were classified as benign group, and the remaining 17 SA associated with malignant lesions were classified as malignant group. On ABVS, 40 mass lesions, their heterogeneous echo, not circumscribed margin and coronal convergence signs were statistically significant for malignant lesions (p< .05), but the remain 48 nonmass lesions lack specific sonographic features. On MG, 12 showed negative results, 55 showed with microcalcification, mass, structural distortion, and asymmetric density shadow, of which 11 lesions had the above two signs at the same time, but only microcalcification had statistical difference between the two groups. 35 mass enhanced lesions and 27 nonmass enhanced lesions on MRI, but there were no significant difference between their pathological results. Time signal intensity curves showed no differences, but ADC value <1.10 × 10-3 mm2 /s is more significant in malignant lesions (p< .05). The area under the ROC curve (AUC) of BI-RADS classification of ABVS, MG, and MRI in the diagnosis of malignant lesions were 0.611, 0.474, and 0.751, respectively, and the AUC of the combined diagnosis of the three was 0.761. Mass lesions with heterogeneous echo, not circumscribed margin and coronal convergence sign on ABVS, microcalcification on MG and the ADC value <1.10× 10-3 mm2 /s on MRI are significant signs for SA associated with malignant lesions. The combined diagnosis of the three methods was the highest, and the following were MRI, ABVS, and MG. Therefore, be cognizant of significant characteristics in SA associated with malignancy showed in different imaging examinations can improve the preoperative evaluation of SA and better provide basis for subsequent clinical decision-making.

Evaluation of the differentiation of benign and malignant breast lesions using synthetic relaxometry and the Kaiser score.

ObjectiveTo investigate whether there is added value of quantitative parameters from synthetic magnetic resonance imaging (SyMRI) as a complement to the Kaiser score (KS) to differentiate benign and malignant breast lesions.Materials and methodsIn this single-institution study, 122 patients who underwent breast MRI from March 2020 to May 2021 were retrospectively analyzed. SyMRI and dynamic contrast-enhanced MRI were performed using a 3.0-T system. Two experienced radiologists independently assigned the KS and measured the quantitative values of T1 relaxation time (T1), T2 relaxation time (T2), and proton density (PD) from SyMRI. Pathology was regarded as the gold standard. The diagnostic values were compared using the appropriate statistical tests.ResultsThere were 122 lesions (86 malignant and 36 benign) in 122 women. The T1 value was identified as the only independent factor for the differentiation of malignant and benign lesions. The diagnostic accuracy of incorporating the T1 into the KS protocol (T1+KS) was 95.1% and 92.1% for all lesions (ALL) and The American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS) category 4 lesions, respectively, which was significantly higher than that of either T1 (ALL: 82.8%, P = 0.0001; BI-RADS 4: 78.9%, P = 0.002) or KS (ALL: 90.2%, P = 0.031; BI-RADS 4: 84.2%, P = 0.031) alone. The sensitivity and specificity of T1+KS were also higher than those of the T1 or KS alone. The combined diagnosis could have avoided another 15.6% biopsies compared with using KS alone.ConclusionsIncorporating T1 into the KS protocol improved both the sensitivity and specificity to differentiate benign and malignant breast lesions, thus avoiding unnecessary invasive procedures.

Breast Density and Breast Cancer Screening with Digital Breast Tomosynthesis: A TOSYMA Trial Subanalysis.

Background Digital breast tomosynthesis (DBT) plus synthesized mammography (SM) reduces the diagnostic pitfalls of tissue superimposition, which is a limitation of digital mammography (DM). Purpose To compare the invasive breast cancer detection rate (iCDR) of DBT plus SM versus DM screening for different breast density categories. Materials and Methods An exploratory subanalysis of the TOmosynthesis plus SYnthesized MAmmography (TOSYMA) study, a randomized, controlled, multicenter, parallel-group trial recruited within the German mammography screening program from July 2018 to December 2020. Women aged 50-69 years were randomly assigned (1:1) to DBT plus SM or DM screening examination. Breast density categories A-D were visually assessed according to the Breast Imaging Reporting and Data System Atlas. Exploratory analyses were performed of the iCDR in both study arms and stratified by breast density, and odds ratios and 95% CIs were determined. Results A total of 49 762 women allocated to DBT plus SM and 49 796 allocated to DM (median age, 57 years [IQR, 53-62 years]) were included. In the DM arm, the iCDR was 3.6 per 1000 screening examinations in category A (almost entirely fatty) (16 of 4475 screenings), 4.3 in category B (102 of 23 534 screenings), 6.1 in category C (116 of 19 051 screenings), and 2.3 in category D (extremely dense breasts) (six of 2629 screenings). The iCDR in the DBT plus SM arm was 2.7 per 1000 screening examinations in category A (12 of 4439 screenings), 6.9 in category B (154 of 22 328 screenings), 8.3 in category C (156 of 18 772 screenings), and 8.1 in category D (32 of 3940 screenings). The odds ratio for DM versus DBT plus SM in category D was 3.8 (95% CI: 1.5, 11.1). The invasive cancers detected with DBT plus SM were most often grade 2 tumors; in category C, it was 58% (91 of 156 invasive cancers), and in category D, it was 47% (15 of 32 invasive cancers). Conclusion The TOmosynthesis plus SYnthesized MAmmography trial revealed higher invasive cancer detection rates with digital breast tomosynthesis plus synthesized mammography than digital mammography in dense breasts, relatively and absolutely most marked among women with extremely dense breasts. ClinicalTrials.gov registration no.: NCT03377036 © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Lee and Moy in this issue.

Analysis of the accuracy of ultrasound elastography and BI-RADS classification of breast masses located within the superficial fat layer of the glands

The surrounding tissue of lesions located in the mammary fat layer is mainly fat, not mammary glands. Are the currently used Breast Imaging Reporting and Data System (BI-RADS) classifications and ultrasound elasticity appropriate for such lesions? In the present study, we discuss the value of elastography and BI-RADS classification in the evaluation of masses in the superficial fat layer of the breast. Seventy-five breast masses within the fat layer that met inclusion criteria were included in the study. Using histopathology as the gold standard, we retrospectively analyzed whether the ultrasound elastography and BI-RADS classification results were consistent with the pathological results. Histological analysis showed that 73 tumors were benign and 2 were malignant. According to the BI-RADS classification and treatment principle, 60% (45/75) of the masses were classified into category 4 and require breast biopsy. But only 4.4% (2/45) of these masses were malignant, and 95.6% (43/45) were overtreated. If we consider the masses with well-defined margins and within the fat layer on the surface of the breast glands as likely benign (BI-RADS category 3), the probability of malignancy is 1.4%. This is consistent with the BI-RADS classification probability of malignancy. According to this BI-RADS classification, only 1.3% (1/75) of patients required biopsy. The findings of this study suggest that breast masses located in the fat layer are prone to be classified into category 4 by BI-RADS and thus be subjected to unnecessary biopsies. Ultrasound elastography can easily misdiagnose benign masses as malignant. It is suggested that ultrasound elastography can downgrade the BI-RADS classification, but not upgrade it. It is more reasonable for these breast masses to be classified as BI-RADS 3 for follow-up observation when the boundary is clear.

Automated BI-RADS classification of lesions using pyramid triple deep feature generator technique on breast ultrasound images.

Ultrasound (US) is an important imaging modality used to assess breast lesions for malignant features. In the past decade, many machine learning models have been developed for automated discrimination of breast cancer versus normal on US images, but few have classified the images based on the Breast Imaging Reporting and Data System (BI-RADS) classes. This work aimed to develop a model for classifying US breast lesions using a BI-RADS classification framework with a new multi-class US image dataset. We proposed a deep model that combined a novel pyramid triple deep feature generator (PTDFG) with transfer learning based on three pre-trained networks for creating deep features. Bilinear interpolation was applied to decompose the input image into four images of successively smaller dimensions, constituting a four-level pyramid for downstream feature generation with the pre-trained networks. Neighborhood component analysis was applied to the generated features to select each network's 1,000 most informative features, which were fed to support vector machine classifier for automated classification using a ten-fold cross-validation strategy. Our proposed model was validated using a new US image dataset containing 1,038 images divided into eight BI-RADS classes and histopathological results. We defined three classification schemes: Case 1 involved the classification of all images into eight categories; Case 2, classification of breast US images into five BI-RADS classes; and Case 3, classification of BI-RADS 4 lesions into benign versus malignant classes. Our PTDFG-based transfer learning model attained accuracy rates of 79.29%, 80.42%, and 88.67% for Case 1, Case 2, and Case 3, respectively.

Imaging and Pathological Correlation in Spectrum of Fibrocystic Breast Disease and its Mimics – our Experience

Background: Fibrocystic change (FCC) of the breast is one of the most common benign breast diseases commonly observed between 20-50 years, with a peak in the perimenopausal age group. Patients present with various symptoms such as lump in the breast, mastalgia (commonly related to the menstrual cycles) or nipple discharge. Materials and Methods: In our retrospective study, which included 172 patients, the imaging findings were observed by ultrasound and X-ray mammogram. Based on the Breast Imaging Reporting and Data System (BI-RADS) guidelines given by the American college of Radiology (ACR), our imaging findings were classified as BI-RADS 2 in benign lesions, complicated cysts were classified as probably benign - BI-RADS 3. Indeterminate findings were classified as suspicious lesions and BI-RADS 4a/b/c. Imaging and histopathological correlation was performed. Results: Ultrasound findings revealed diffuse/ bilateral abnormalities with the most common finding being simple cysts followed in descending order by complicated cysts, clustered cysts, complex solid cystic masses, solid lesions, duct ectasia, and intraductal lesions. Mammogram showed dense (type C or D) fibro glandular pattern obscuring the lesions, followed by well-defined / partly obscured opacities. Simple cysts and complicated cysts showed predominant features of cyst formation on HPE. Atypical hyperplasia was seen in ductal and complex solid cystic mass lesions. Conclusion: It is essential for radiologists to be familiar with imaging and pathological findings of fibrocystic disease of the breast for further workup and management.