Search

We investigate the predictive value of a comprehensive model based on preoperative ultrasound radiomics, deep learning, and clinical features for pathological complete response (pCR) after neoadjuvant chemotherapy (NAC) for the breast cancer. We enrolled 155 patients with pathologically confirmed breast cancer who underwent NAC. The patients were randomly divided into the training set and the validation set in the ratio of 7:3. The deep learning and radiomics features of pre-treatment ultrasound images were extracted, and the random forest recursive elimination algorithm and the least absolute shrinkage and selection operator were used for feature screening and DL-Score and Rad-Score construction. According to multifactorial logistic regression, independent clinical predictors, DL-Score, and Rad-Score were selected to construct the comprehensive prediction model DLRC. The performance of the model was evaluated in terms of its predictive effect, and clinical practicability. Compared to the clinical, radiomics (Rad-Score), and deep learning (DL-Score) models, the DLRC accurately predicted the pCR status, with an area under the curve (AUC) of 0.937 (95%CI: 0.895-0.970) in the training set and 0.914 (95%CI: 0.838-0.973) in the validation set. Moreover, decision curve analysis confirmed that the DLRC had the highest clinical value among all models. The comprehensive model DLRC based on ultrasound radiomics, deep learning, and clinical features can effectively and accurately predict the pCR status of breast cancer after NAC, which is conducive to assisting clinical personalized diagnosis and treatment plan., Competing Interests: Declaration of Conflicting InterestsThe author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Breast ultrasound utilizes various scanning techniques to acquire optimal images for diagnostic evaluation. During interventional procedures, such as ultrasound-guided biopsies or preoperative localizations, knowledgeable and purposeful scanning adjustments are critical for successfully identifying the targeted mass or biopsy marker or clip. While most ultrasound scanning parameters are similar across different vendors, detailed descriptions specifically addressing the scanning parameters-often referred to as "knobology"- for breast ultrasound is at best limited in the literature. This review highlights ten key operator-controlled adjustments (including transducer selection, beam focusing, power, depth, gain and time gain compensation, harmonic imaging, spatial compounding, dynamic range, beam steering, and color Doppler) that significantly influence image quality in breast ultrasound. Each adjustment is accompanied by an "In practice" section providing examples and practical tips on implementation. The last topic discusses color Doppler which is generally used in breast ultrasound for evaluating the vascularity of a finding. Color Doppler, or more specifically, color Doppler twinkling, can be leveraged as a technique to detect certain breast biopsy markers that are challenging to detect by conventional B-mode ultrasound. While the cause of color Doppler twinkling is still under active investigation, twinkling is a clinically well-known, compelling ultrasound feature typically described with kidney stones. A step-by-step guide on how to use color Doppler twinkling to detect these markers is provided., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tbcr.amegroups.org/article/view/10.21037/tbcr-24-30/coif). C.U.L. serves as an unpaid editorial board member of Translational Breast Cancer Research from October 2023 to September 2025. C.U.L. and M.W.U. are co-PIs on a National Institute of Biomedical Imaging and Bioengineering (NIBIB) grant (R01 EB033008) to the Mayo Clinic and on an internal Mayo Clinic Discovery Translation Program award. Patents are pending for non-metallic ultrasound detectable markers as well as Doppler ultrasound twinkling technologies. G.K.H. was supported by NIBIB grant (R01 EB033008) to the Mayo Clinic and by an internal Mayo Clinic Discovery Translation Program award. The other authors have no other conflicts of interest to declare., (2024 AME Publishing Company. All rights reserved.)

Objective. Ultrasound is the primary screening test for breast cancer. However, providing an interpretable auxiliary diagnosis of breast lesions is a challenging task. This study aims to develop an interpretable auxiliary diagnostic method to enhance usability in human-machine collaborative diagnosis. Approach. To address this issue, this study proposes the deep multi-stage reasoning method (DMSRM), which provides individual and overall breast imaging-reporting and data system (BI-RADS) assessment categories for breast lesions. In the first stage of the DMSRM, the individual BI-RADS assessment network (IBRANet) is designed to capture lesion features from breast ultrasound images. IBRANet performs individual BI-RADS assessments of breast lesions using ultrasound images, focusing on specific features such as margin, contour, echogenicity, calcification, and vascularity. In the second stage, evidence reasoning (ER) is employed to achieve uncertain information fusion and reach an overall BI-RADS assessment of the breast lesions. Main results. To evaluate the performance of DMSRM at each stage, two test sets are utilized: the first for individual BI-RADS assessment, containing 4322 ultrasound images; the second for overall BI-RADS assessment, containing 175 sets of ultrasound image pairs. In the individual BI-RADS assessment of margin, contour, echogenicity, calcification, and vascularity, IBRANet achieves accuracies of 0.9491, 0.9466, 0.9293, 0.9234, and 0.9625, respectively. In the overall BI-RADS assessment of lesions, the ER achieves an accuracy of 0.8502. Compared to independent diagnosis, the human-machine collaborative diagnosis results of three radiologists show increases in positive predictive value by 0.0158, 0.0427, and 0.0401, in sensitivity by 0.0400, 0.0600 and 0.0434, and in area under the curve by 0.0344, 0.0468, and 0.0255. Significance. This study proposes a DMSRM that enhances the transparency of the diagnostic reasoning process. Results indicate that DMSRM exhibits robust BI-RADS assessment capabilities and provides an interpretable reasoning process that better suits clinical needs., (© 2024 Institute of Physics and Engineering in Medicine. All rights, including for text and data mining, AI training, and similar technologies, are reserved.)

Rationale and Objectives: The combination of functional biologic data and imaging appearance has the potential to add diagnostic information to help the radiologist evaluate breast masses in an efficient, effective, and cost-conscious manner. This is the first clinical evaluation of the Gen 2(Model 9100, 8101) Imagio® System to assess image quality with both the stand-alone internal ultrasound (IUS), ultrasound-only transducer, and the Optoacoustic/Ultrasound (OA/US) duplex probe (1,2). This study assesses palpable and non-palpable breast abnormalities in patients who are referred for diagnostic breast ultrasound work-up. This study is intended to confirm the clinical acceptability of modifications made to the Imagio® System ultrasound component following Premarket Approval (PMA) of the Imagio® Gen 1 version., Materials and Methods: This prospective, single-arm, non-randomized study included 38 patients presenting with a palpable lump and/or imaging abnormality detected at a single investigational site. Each patient had the breast, and if indicated, the axillary lymph nodes imaged with the Gen 2 Imagio® system., Results: For patients with SenoGram®-predicted Likelihood of Malignancy (LOM) and pathology available (N = 23), observed sensitivity was 100.0% (9/9) with a confidence interval of (66.4%, 100.0%), using a SenoGram®-predicted False Negative Rate (FNR) cut-off of ≤ 2%. Observed specificity was 64.3% (9/14) (Confidence Interval: 35.1%, 87.2%), using a SenoGram®-predicted FNR cut-off of ≤ 2%. At 98% fixed sensitivity, the specificity (fSp) for OA/US + SG was 100.0% while it was 0.0% for IUS. The absolute gain in fSp was 100.0%., Conclusion: Combining structure with morphology can increase specificity without decreasing sensitivity in a real-world setting., Competing Interests: Declaration of Competing Interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Kenneth A Kist reports a relationship with Seno Medical Instruments that includes: consulting or advisory. Dr. Pamela Otto is on the Advisory Board for Seno. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.)

This study addresses the challenge of precise breast tumor segmentation in ultrasound images, crucial for effective Computer-Aided Diagnosis (CAD) in breast cancer. We introduce CBAM-RIUnet, a deep learning (DL) model for automated breast tumor segmentation in breast ultrasound (BUS) images. The model, featuring an efficient convolutional block attention module residual inception Unet, outperforms existing models, particularly excelling in Dice and IoU scores. CBAM-RIUnet follows the Unet structure with a residual inception depth-wise separable convolution, and incorporates a convolutional block attention module (CBAM) to eliminate irrelevant features and focus on the region of interest. Evaluation under three scenarios, including enhanced breast ultrasound (EBUS) and test-time augmentation (TTA), demonstrates impressive results. CBAM-RIUnet achieves Dice and IoU scores of 89.38% and 88.71%, respectively, showcasing significant improvements compared to state-of-the-art DL techniques. In conclusion, CBAM-RIUnet presents a highly effective and simplified DL model for breast tumor segmentation in BUS imaging., Competing Interests: Declaration of Conflicting InterestsThe author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Core needle biopsy is a common diagnostic procedure in breast cancer patients, but it can occasionally lead to serious complications. We report a rare case of pseudoaneurysm formation following a core needle biopsy in a 54-year-old female patient diagnosed with breast cancer. Despite the routine nature of the procedure, the patient developed a palpable mass at the biopsy site, which prompted further diagnostic imaging and interventions. The pseudoaneurysm was effectively treated using a percutaneous approach with ultrasound-guided thrombin injection, demonstrating a minimally invasive solution that promptly addressed the complication without the need for surgical intervention. This case highlights the critical importance of detecting complications early in the biopsy process, as they have significant implications for disease staging and treatment initiation. It also underscores the importance of being prepared for immediate intervention in case of biopsy-related complications like pseudoaneurysms, to prevent severe consequences., (© 2024 The Authors. Published by Elsevier Inc. on behalf of University of Washington.)

Invasive lobular carcinoma (ILC) is a rare type of male breast cancer, representing about 1% of cases. It often presents with a palpable mass and sometimes nipple changes, but is usually diagnosed late. ILC is more likely to be estrogen and progesterone receptor-positive and usually HER-2 negative. Its diffuse growth pattern makes it difficult to detect with imaging. Treatment typically follows protocols for female breast cancer, including surgery, chemotherapy, and hormone therapy, with tamoxifen being commonly used. Further research is needed to better understand its pathogenesis and to develop more effective, tailored treatments., (© 2024 The Authors. Published by Elsevier Inc. on behalf of University of Washington.)

Breast cancer is the most prevalent cancer among women worldwide. B-mode ultrasound (US) is essential for early detection, offering high sensitivity and specificity without radiation exposure. This study introduces a semi-automatic method to streamline breast US report generation, aiming to reduce the burden on radiologists. Our method synthesizes comprehensive breast US reports by combining the extracted information from radiologists' annotations during routine screenings with the analysis results from deep learning algorithms on multimodal US images. Key modules in our method include image classification using visual features (ICVF), type classification via deep learning (TCDL), and automatic report structuring and compilation (ARSC). Experiments showed that the proposed method reduced the average report generation time to 3.8 min compared to manual processes, even when using relatively low-spec hardware. Generated reports perfectly matched ground truth reports for suspicious masses without a single failure on our evaluation datasets. Additionally, the deep-learning-based algorithm, utilizing DenseNet-121 as its core model, achieved an overall accuracy of 0.865, precision of 0.868, recall of 0.847, F1-score of 0.856, and area under the receiver operating characteristics of 0.92 in classifying tissue stiffness in breast US shear-wave elastography (SWE-mode) images. These improvements not only streamline the report generation process but also allow radiologists to dedicate more time and focus on patient care, ultimately enhancing clinical outcomes and patient satisfaction.

Purpose: The Ugandan Ministry of Health adopted BI-RADS as standard of care in 2016. The authors performed a medical audit of breast ultrasound practices at four tertiary-level hospitals to assess interpretive performance. The authors also determined the effect of a low-cost navigation program linking breast imaging and pathology on the percentage of patients completing diagnostic care., Methods: The authors retrieved 966 consecutive diagnostic breast ultrasound reports, with complete data, for studies performed on women aged >18 years presenting with symptoms of breast cancer between 2018 and 2020 from participating hospitals. Ultrasound results were linked to tumor registries and patient follow-up. A medical audit was performed according to the ACR's BI-RADS Atlas, fifth edition, and results were compared with those of a prior audit performed in 2013. At Mulago Hospital, an intervention was piloted on the basis of patient navigation, cost sharing, and same-day imaging, tissue sampling, and pathology., Results: In total, 888 breast ultrasound examinations (91.9%) were eligible for inclusion. Compared with 2013, the postintervention cancer detection rate increased from 38 to 148.7 cancers per 1,000 examinations, positive predictive value 2 from 29.6% to 48.9%, and positive predictive value 3 from 62.7% to 79.9%. Specificity decreased from 90.5% to 87.7% and sensitivity from 92.3% to 81.1%. The mean time from tissue sampling to receipt of a diagnosis decreased from 60 to 7 days. The intervention increased the percentage of patients completing diagnostic care from 0% to 100%., Conclusions: Efforts to establish a culture of continuous quality improvement in breast ultrasound require robust data collection that links imaging results to pathology and patient follow-up. Interpretive performance met BI-RADS benchmarks for palpable masses, except sensitivity. This resource-appropriate strategy linking imaging, tissue sampling, and pathology interpretation decreased time to diagnosis and rates of loss to follow-up and improved the precision of the audit., (Copyright © 2024 American College of Radiology. Published by Elsevier Inc. All rights reserved.)

Automatic breast ultrasound image segmentation plays an important role in medical image processing. However, current methods for breast ultrasound segmentation suffer from high computational complexity and large model parameters, particularly when dealing with complex images. In this paper, we take the Unext network as a basis and utilize its encoder-decoder features. And taking inspiration from the mechanisms of cellular apoptosis and division, we design apoptosis and division algorithms to improve model performance. We propose a novel segmentation model which integrates the division and apoptosis algorithms and introduces spatial and channel convolution blocks into the model. Our proposed model not only improves the segmentation performance of breast ultrasound tumors, but also reduces the model parameters and computational resource consumption time. The model was evaluated on the breast ultrasound image dataset and our collected dataset. The experiments show that the SC-Unext model achieved Dice scores of 75.29% and accuracy of 97.09% on the BUSI dataset, and on the collected dataset, it reached Dice scores of 90.62% and accuracy of 98.37%. Meanwhile, we conducted a comparison of the model's inference speed on CPUs to verify its efficiency in resource-constrained environments. The results indicated that the SC-Unext model achieved an inference speed of 92.72 ms per instance on devices equipped only with CPUs. The model's number of parameters and computational resource consumption are 1.46M and 2.13 GFlops, respectively, which are lower compared to other network models. Due to its lightweight nature, the model holds significant value for various practical applications in the medical field., (© 2024. The Author(s) under exclusive licence to Society for Imaging Informatics in Medicine.)

Objective: Evaluate surgical utilization of SCOUT reflectors placed at breast biopsy., Methods: Consent was waived for this retrospective IRB-approved, HIPAA-compliant study. Breast biopsy examinations that reported the term "SCOUT" between January 2021 and June 2022 were identified using an institutional search engine. Cases were included if a SCOUT reflector was placed at time of breast biopsy and excluded if lesion pathology was already known. Analysis was performed at the lesion level. A multivariate-regression analysis evaluated 6 variables with potential impact on SCOUT utilization., Results: One hundred twenty-one lesions in 112 patients met inclusion criteria. Biopsy yielded 93% (113/121) malignant, 3% (4/121) elevated risk, 2% (2/121) benign-discordant, and 2% (2/121) benign-concordant results. Two cases lost to follow-up were excluded. SCOUT reflectors were utilized for lumpectomy (58%, 69/119 lesions) and excisional biopsy (6%, 7/119 lesions). SCOUTs were not utilized due to mastectomy (23%, 27/119), subsequent wire localization (2%, 2/119), and nonsurgical cases (12%, 14/119). Reflector placement utilization was 52% higher for findings less than 3.5 cm in size (P <.001), 33% higher in patients without prior treated breast cancer (P = .012), and 19% higher in patients with no suspicious ipsilateral lymph node (P = .048)., Conclusion: SCOUT reflector placement at time of biopsy was utilized for surgery 64% (76/119) of the time, although most (98%, 119/121) biopsies were malignant, elevated risk, or benign-discordant. Factors increasing reflector utilization include smaller lesion size, no suspicious ipsilateral lymph node, and no prior treated breast cancer., (© Society of Breast Imaging 2024. All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.)

Interoperator variability in the reproducibility of breast lesions found by handheld ultrasound (HHUS) can significantly interfere with clinical care. This study analyzed the features associated with breast mass position differences during HHUS. The ability of operators to reproduce the position of small masses and the time required to generate annotations with and without a computer-assisted scanning device (DEVICE) were also evaluated. This prospective study included 28 patients with 34 benign or probably benign small breast masses. Two operators generated manual and automated position annotations for each mass. The probe and body positions were systematically varied during scanning with the DEVICE, and the features describing mass movement were used in three logistic regression models trained to discriminate small from large breast mass displacements (cutoff: 10 mm). All models successfully discriminated small from large breast mass displacements (areas under the curve: 0.78 to 0.82). The interoperator localization precision was 6.6 ± 2.8 mm with DEVICE guidance and 19.9 ± 16.1 mm with manual annotations. Computer-assisted scanning reduced the time to annotate and reidentify a mass by 33 and 46 s on average, respectively. The results demonstrated that breast mass location reproducibility and exam efficiency improved by controlling operator actionable features with computer-assisted HHUS.

Purpose: To establish a reliable machine learning model to predict malignancy in breast lesions identified by ultrasound (US) and optimize the negative predictive value to minimize unnecessary biopsies., Methods: We included clinical and ultrasonographic attributes from 1526 breast lesions classified as BI-RADS 3, 4a, 4b, 4c, 5, and 6 that underwent US-guided breast biopsy in four institutions. We selected the most informative attributes to train nine machine learning models, ensemble models and models with tuned threshold to make inferences about the diagnosis of BI-RADS 4a and 4b lesions (validation dataset). We tested the performance of the final model with 403 new suspicious lesions., Results: The most informative attributes were shape, margin, orientation and size of the lesions, the resistance index of the internal vessel, the age of the patient and the presence of a palpable lump. The highest mean negative predictive value (NPV) was achieved with the K-Nearest Neighbors algorithm (97.9%). Making ensembles did not improve the performance. Tuning the threshold did improve the performance of the models and we chose the algorithm XGBoost with the tuned threshold as the final one. The tested performance of the final model was: NPV 98.1%, false negative 1.9%, positive predictive value 77.1%, false positive 22.9%. Applying this final model, we would have missed 2 of the 231 malignant lesions of the test dataset (0.8%)., Conclusion: Machine learning can help physicians predict malignancy in suspicious breast lesions identified by the US. Our final model would be able to avoid 60.4% of the biopsies in benign lesions missing less than 1% of the cancer cases., (© 2024. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.)

Systemic lupus erythematosus (SLE) and sarcoidosis are two of the most well-recognized, chronically diagnosed conditions in the United States, with a plethora of known multisystem manifestations. With regard to breast pathology, lupus mastitis is a relatively uncommon manifestation of SLE, commonly involving both the mammary gland and subcutaneous soft tissues of the breast. Sarcoidosis in the breast is a similarly, exceedingly rare manifestation of this multi-system disorder, classically presenting with non-caseating granulomas. Both present with non-specific mammographic and sonographic features. We present a 62-year-old female with known diagnosis of discoid lupus and Graves' disease who presented initially with an abnormal screening mammogram, ultimately undergoing mammographic work-up and subsequent biopsy demonstrating lupus mastitis, including vasculitis, panniculitis, and fibrosis with chronic inflammation. The patient was also found to have small non-caseating granulomas, some in a perivascular distribution, classically seen in sarcoidosis. Given the rarity of both manifestations, our case explores the coexistence of these autoimmune processes and this atypical presentation., Competing Interests: Conflict of Interest: No conflict of interest was declared by the authors., (©Copyright 2024 by the Turkish Federation of Breast Diseases Societies / European Journal of Breast Health published by Galenos Publishing House.)

Breast cancer is the most common cancer in women. Ultrasound is one of the most used techniques for diagnosis, but an expert in the field is necessary to interpret the test. Computer-aided diagnosis (CAD) systems aim to help physicians during this process. Experts use the Breast Imaging-Reporting and Data System (BI-RADS) to describe tumors according to several features (shape, margin, orientation...) and estimate their malignancy, with a common language. To aid in tumor diagnosis with BI-RADS explanations, this paper presents a deep neural network for tumor detection, description, and classification. An expert radiologist described with BI-RADS terms 749 nodules taken from public datasets. The YOLO detection algorithm is used to obtain Regions of Interest (ROIs), and then a model, based on a multi-class classification architecture, receives as input each ROI and outputs the BI-RADS descriptors, the BI-RADS classification (with 6 categories), and a Boolean classification of malignancy. Six hundred of the nodules were used for 10-fold cross-validation (CV) and 149 for testing. The accuracy of this model was compared with state-of-the-art CNNs for the same task. This model outperforms plain classifiers in the agreement with the expert (Cohen's kappa), with a mean over the descriptors of 0.58 in CV and 0.64 in testing, while the second best model yielded kappas of 0.55 and 0.59, respectively. Adding YOLO to the model significantly enhances the performance (0.16 in CV and 0.09 in testing). More importantly, training the model with BI-RADS descriptors enables the explainability of the Boolean malignancy classification without reducing accuracy., (© 2024. The Author(s).)

Purpose: Preventing positive margins is essential for ensuring favorable patient outcomes following breast-conserving surgery (BCS). Deep learning has the potential to enable this by automatically contouring the tumor and guiding resection in real time. However, evaluation of such models with respect to pathology outcomes is necessary for their successful translation into clinical practice., Methods: Sixteen deep learning models based on established architectures in the literature are trained on 7318 ultrasound images from 33 patients. Models are ranked by an expert based on their contours generated from images in our test set. Generated contours from each model are also analyzed using recorded cautery trajectories of five navigated BCS cases to predict margin status. Predicted margins are compared with pathology reports., Results: The best-performing model using both quantitative evaluation and our visual ranking framework achieved a mean Dice score of 0.959. Quantitative metrics are positively associated with expert visual rankings. However, the predictive value of generated contours was limited with a sensitivity of 0.750 and a specificity of 0.433 when tested against pathology reports., Conclusion: We present a clinical evaluation of deep learning models trained for intraoperative tumor segmentation in breast-conserving surgery. We demonstrate that automatic contouring is limited in predicting pathology margins despite achieving high performance on quantitative metrics., (© 2024. CARS.)

Purpose: Primary Mucosa-associated lymphoid tissue (MALT) lymphoma is a rare diagnosis in the breast, and clinical diagnosis based on radiological features is often challenging. This study aimed to evaluate the clinicopathological, and radiological characteristics of the patients diagnosed with primary breast MALT lymphoma., Methods: This study examined 18 cases of primary MALT lymphoma of the breast diagnosed at a single tertiary center between January 2002 to December 2020. Medical charts, radiological imaging and original pathology slides were reviewed for each case., Results: All cases were female (gender assigned at birth) and presented with a palpable mass or an incidental imaging finding. Imaging presentation ranged from mammographic asymmetries, circumscribed masses, and ultrasound masses lacking suspicious features. Seventeen cases were biopsied under ultrasound; one received a diagnostic excision biopsy. Microscopic examination of the breast specimens demonstrated atypical small lymphocyte infiltration with plasmacytoid differentiation and rare lymphoepithelial lesions. Immunohistochemistry was performed in all cases and established the diagnosis. Most patients were treated with radiotherapy, and only three were treated with chemotherapy. The median follow-up period was 4 years and 7.5 months, and all patients were alive at the last follow-up., Conclusion: Primary MALT breast lymphomas are usually indolent and non-systemic, and local radiotherapy may effectively alleviate local symptoms. Radiological findings show overlap with benign morphological features, which can delay the diagnosis of this unusual etiology. Although further studies involving a larger cohort could help establish the clinical and radiological characteristics of primary breast MALT lymphomas, pathology remains the primary method of diagnosis., Trial Registration Number: University Health Network Ethics Committee (CAPCR/UHN REB number 19-5844), retrospectively registered., (© 2024. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.)

Background: Accurate classification of breast nodules into benign and malignant types is critical for the successful treatment of breast cancer. Traditional methods rely on subjective interpretation, which can potentially lead to diagnostic errors. Artificial intelligence (AI)-based methods using the quantitative morphological analysis of ultrasound images have been explored for the automated and reliable classification of breast cancer. This study aimed to investigate the effectiveness of AI-based approaches for improving diagnostic accuracy and patient outcomes., Methods: In this study, a quantitative analysis approach was adopted, with a focus on five critical features for evaluation: degree of boundary regularity, clarity of boundaries, echo intensity, and uniformity of echoes. Furthermore, the classification results were assessed using five machine learning methods: logistic regression (LR), support vector machine (SVM), decision tree (DT), naive Bayes, and K-nearest neighbor (KNN). Based on these assessments, a multifeature combined prediction model was established., Results: We evaluated the performance of our classification model by quantifying various features of the ultrasound images and using the area under the receiver operating characteristic (ROC) curve (AUC). The moment of inertia achieved an AUC value of 0.793, while the variance and mean of breast nodule areas achieved AUC values of 0.725 and 0.772, respectively. The convexity and concavity achieved AUC values of 0.988 and 0.987, respectively. Additionally, we conducted a joint analysis of multiple features after normalization, achieving a recall value of 0.98, which surpasses most medical evaluation indexes on the market. To ensure experimental rigor, we conducted cross-validation experiments, which yielded no significant differences among the classifiers under 5-, 8-, and 10-fold cross-validation (P>0.05)., Conclusions: The quantitative analysis can accurately differentiate between benign and malignant breast nodules., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-23-1652/coif). The authors have no conflicts of interest to declare., (2024 Quantitative Imaging in Medicine and Surgery. All rights reserved.)

Background  Radial breast ultrasound scanning (r-US) and commonly used meander-like ultrasound scanning (m-US) have recently been shown to be equally sensitive and specific with regard to the detection of breast malignancies. As patient satisfaction has a strong influence on patient compliance and thus on the quality of health care, we compare here the two US scanning techniques with regard to patient comfort during breast ultrasound (BUS) and analyze whether the patient has a preference for either scanning technique. Materials and Methods  Symptomatic and asymptomatic women underwent both m-US and r-US scanning by two different examiners. Patient comfort and preference were assessed using a visual analog scale-based (VAS) questionnaire and were compared using a Mann-Whitney U test. Results  Analysis of 422 VAS-based questionnaires showed that perceived comfort with r-US (r-VAS 8 cm, IQR [5.3, 9.1]) was significantly higher compared to m-US (m-VAS 5.6 cm, IQR [5.2, 7.4]) (p < 0.001). 53.8% of patients had no preference, 44.3% of patients clearly preferred r-US, whereas only 1.9% of patients preferred m-US. Conclusion: Patients experience a higher level of comfort with r-US and favor r-US over m-US. As the diagnostic accuracy of r-US has been shown to be comparable to that of m-US and the time required for examination is shorter, a switch from m-US to r-US in routine clinical practice might be beneficial. R-US offers considerable potential to positively affect patient compliance but also to save examination time and thus costs., Competing Interests: Conflict of Interest The authors declare that they have no conflict of interest., (The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial-License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. (https://creativecommons.org/licenses/by-nc-nd/4.0/).)

Background: Owing to its high sensitivity, as concluded in the Breast UltraSound Trial (BUST), targeted ultrasound (US) now seems a promising accurate stand-alone modality for diagnostic evaluation of breast complaints. This approach implies omission of bilateral digital breast tomosynthesis (DBT) in women with clearly benign US findings. Within BUST, radiologists started with US followed by DBT. This side-study investigates women's experiences with DBT, their main motivation to undergo diagnostic imaging, and their view on US as a stand-alone modality. Methods: A subset of BUST participants completed a questionnaire on their DBT experiences, reason for undergoing diagnostic assessment, and view on US-only diagnostics. Responses were analyzed with descriptive statistics and logistic regression analyses. Results: In total, 778 of 838 women (response rate 92.8%) were included ( M  = 47, SD  = 11.16). Of them, 16.8% reported no burden of DBT, 33.5% slight burden, 31.0% moderate, and 12.7% severe burden. Furthermore, 13% reported no pain, 35.3% slight pain, 33.2% moderate, and 11.3% severe pain. Moreover, 88.3% indicated that the most important reason for breast assessment was explanation of their complaint and to rule out breast cancer, whereas 3.2% wanted to "check" both breasts. And 82.4% reported satisfaction with US only in case of a nonmalignancy. Conclusions: Our study shows that most women in the diagnostic setting experience at least slight-to-moderate DBT-related burden and pain, and that explanation for their symptoms is their main interest. Also, the majority report satisfaction with US only in case of nonmalignant findings. However, exploration of women's perspectives outside this study is needed as our participants all underwent both examinations.

Purpose: Computer-aided diagnosis (CAD) systems on breast ultrasound (BUS) aim to increase the efficiency and effectiveness of breast screening, helping specialists to detect and classify breast lesions. CAD system development requires a set of annotated images, including lesion segmentation, biopsy results to specify benign and malignant cases, and BI-RADS categories to indicate the likelihood of malignancy. Besides, standardized partitions of training, validation, and test sets promote reproducibility and fair comparisons between different approaches. Thus, we present a publicly available BUS dataset whose novelty is the substantial increment of cases with the above-mentioned annotations and the inclusion of standardized partitions to objectively assess and compare CAD systems., Acquisition and Validation Methods: The BUS dataset comprises 1875 anonymized images from 1064 female patients acquired via four ultrasound scanners during systematic studies at the National Institute of Cancer (Rio de Janeiro, Brazil). The dataset includes biopsy-proven tumors divided into 722 benign and 342 malignant cases. Besides, a senior ultrasonographer performed a BI-RADS assessment in categories 2 to 5. Additionally, the ultrasonographer manually outlined the breast lesions to obtain ground truth segmentations. Furthermore, 5- and 10-fold cross-validation partitions are provided to standardize the training and test sets to evaluate and reproduce CAD systems. Finally, to validate the utility of the BUS dataset, an evaluation framework is implemented to assess the performance of deep neural networks for segmenting and classifying breast lesions., Data Format and Usage Notes: The BUS dataset is publicly available for academic and research purposes through an open-access repository under the name BUS-BRA: A Breast Ultrasound Dataset for Assessing CAD Systems. BUS images and reference segmentations are saved in Portable Network Graphic (PNG) format files, and the dataset information is stored in separate Comma-Separated Value (CSV) files., Potential Applications: The BUS-BRA dataset can be used to develop and assess artificial intelligence-based lesion detection and segmentation methods, and the classification of BUS images into pathological classes and BI-RADS categories. Other potential applications include developing image processing methods like despeckle filtering and contrast enhancement methods to improve image quality and feature engineering for image description., (© 2023 American Association of Physicists in Medicine.)

Background: While breast ultrasound (US) is a useful tool for diagnosing breast masses, it can entail false-positive biopsy results because of some overlapping features between benign and malignant breast masses and subjective interpretation., Purpose: To evaluate the performance of conductivity imaging for reducing false-positive biopsy results related to breast US, as compared to diffusion-weighted imaging (DWI) and abbreviated MRI consisting of one pre- and one post-contrast T1-weighted imaging., Study Type: Prospective., Subjects: Seventy-nine women (median age, 44 years) with 86 Breast Imaging Reporting and Data System (BI-RADS) category 4 masses as detected by breast US., Field Strength/sequence: 3-T, T2-weighted turbo spin echo sequence, DWI, and abbreviated contrast-enhanced MRI (T1-weighted gradient echo sequence)., Assessment: US-guided biopsy (reference standard) was obtained on the same day as MRI. The maximum and mean conductivity parameters from whole and single regions of interest (ROIs) were measured. Apparent diffusion coefficient (ADC) values were obtained from an area with the lowest signal within a lesion on the ADC map. The performance of conductivity, ADC, and abbreviated MRI for reducing false-positive biopsies was evaluated using the following criteria: lowest conductivity and highest ADC values among malignant breast lesions and BI-RADS categories 2 or 3 on abbreviated MRI., Statistical Tests: One conductivity parameter with the maximum area under the curve (AUC) from receiver operating characteristics was selected. A P-value <0.05 was considered statistically significant., Results: US-guided biopsy revealed 65 benign lesions and 21 malignant lesions. The mean conductivity parameter of the single ROI method was selected (AUC = 0.74). Considering conductivity (≤0.10 S/m), ADC (≥1.60 × 10 -3  mm 2 /sec), and BI-RADS categories 2 or 3 reduced false-positive biopsies by 23% (15 of 65), 38% (25 of 65), and 43% (28 of 65), respectively, without missing malignant lesions., Data Conclusion: Conductivity imaging may show lower performance than DWI and abbreviated MRI in reducing unnecessary biopsies., Evidence Level: 2 TECHNICAL EFFICACY: Stage 2., (© 2023 International Society for Magnetic Resonance in Medicine.)

Objective: To determine breast radiologists' confidence in detecting invasive lobular carcinoma (ILC) on mammography and the perceived need for additional imaging in screening and preoperative settings., Methods: A 16-item anonymized survey was developed, and IRB exemption obtained, by the Society of Breast Imaging (SBI) Patient Care and Delivery Committee and the Lobular Breast Cancer Alliance. The survey was emailed to 2946 radiologist SBI members on February 15, 2023. The survey recorded demographics, perceived modality-specific sensitivity for ILC to the nearest decile, and opinions on diagnosing ILC in screening and staging imaging. Five-point Likert scales were used (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree)., Results: Response rate was 12.4% (366/2946). Perceived median (interquartile range) modality-specific sensitivities for ILC were MRI 90% (80-90), contrast-enhanced mammography 80% (70-90), molecular breast imaging 80% (60-90), digital breast tomosynthesis 70% (60-80), US 60% (50-80), and 2D mammography 50% (30-60). Only 25% (85/340) respondents were confident in detecting ILC on screening mammography in dense breasts, while 67% (229/343) were confident if breasts were nondense. Most agreed that supplemental screening is needed to detect ILC in women with dense breasts (272/344, 79%) or a personal history of ILC (248/341, 73%), with 34% (118/334) indicating that supplemental screening would also benefit women with nondense breasts. Most agreed that additional imaging is needed to evaluate extent of disease in women with newly diagnosed ILC, regardless of breast density (dense 320/329, 97%; nondense 263/329, 80%)., Conclusion: Most breast radiologists felt that additional imaging beyond mammography is needed to more confidently screen for and stage ILC., (© Society of Breast Imaging 2024.)

Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical imaging. Due to the poor quality of US images and the varying specifications of US machines, segmentation and classification of abnormalities present difficulties even for trained radiologists. The paper aims to introduce a novel AI-based hybrid model for US segmentation that offers high accuracy, requires relatively smaller datasets, and is capable of handling previously unseen data. The software can be used for diagnostics and the US-guided biopsies. A unique and robust hybrid approach that combines deep learning (DL) and multi-agent artificial life (AL) has been introduced. The algorithms are verified on three US datasets. The method outperforms 14 selected state-of-the-art algorithms applied to US images characterized by complex geometry and high level of noise. The paper offers an original classification of the images and tests to analyze the limits of the DL. The model has been trained and verified on 1264 ultrasound images. The images are in the JPEG and PNG formats. The age of the patients ranges from 22 to 73 years. The 14 benchmark algorithms include deformable shapes, edge linking, superpixels, machine learning, and DL methods. The tests use eight-region shape- and contour-based evaluation metrics. The proposed method (DL-AL) produces excellent results in terms of the dice coefficient (region) and the relative Hausdorff distance H 3 (contour-based) as follows: the easiest image complexity level, Dice = 0.96 and H 3  = 0.26; the medium complexity level, Dice = 0.91 and H 3  = 0.82; and the hardest complexity level, Dice = 0.90 and H 3  = 0.84. All other metrics follow the same pattern. The DL-AL outperforms the second best (Unet-based) method by 10-20%. The method has been also tested by a series of unconventional tests. The model was trained on low complexity images and applied to the entire set of images. These results are summarized below. (1) Only the low complexity images have been used for training (68% unknown images): Dice = 0.80 and H 3  = 2.01. (2) The low and the medium complexity images have been used for training (51% unknown images): Dice = 0.86 and H 3  = 1.32. (3) The low, medium, and hard complexity images have been used for training (35% unknown images): Dice = 0.92 and H 3  = 0.76. These tests show a significant advantage of DL-AL over 30%. A video demo illustrating the algorithm is at http://tinyurl.com/mr4ah687 ., (© 2024. International Federation for Medical and Biological Engineering.)

Objective: To compare the medical image interpretation's time between the conventional and automated methods of breast ultrasound in patients with breast lesions. Secondarily, to evaluate the agreement between the two methods and interobservers., Methods: This is a cross-sectional study with prospective data collection. The agreement's degrees were established in relation to the breast lesions's ultrasound descriptors. To determine the accuracy of each method, a biopsy of suspicious lesions was performed, considering the histopathological result as the diagnostic gold standard., Results: We evaluated 27 women. Conventional ultrasound used an average medical time of 10.77 minutes (± 2.55) greater than the average of 7.38 minutes (± 2.06) for automated ultrasound (p<0.001). The degrees of agreement between the methods ranged from 0.75 to 0.95 for researcher 1 and from 0.71 to 0.98 for researcher 2. Among the researchers, the degrees of agreement were between 0.63 and 1 for automated ultrasound and between 0.68 and 1 for conventional ultrasound. The area of the ROC curve for the conventional method was 0.67 (p=0.003) for researcher 1 and 0.72 (p<0.001) for researcher 2. The area of the ROC curve for the automated method was 0. 69 (p=0.001) for researcher 1 and 0.78 (p<0.001) for researcher 2., Conclusion: We observed less time devoted by the physician to automated ultrasound compared to conventional ultrasound, maintaining accuracy. There was substantial or strong to perfect interobserver agreement and substantial or strong to almost perfect agreement between the methods., Competing Interests: Conflicts to interest: none to declare., (© 2024. Federação Brasileira de Ginecologia e Obstetrícia. All rights reserved.)

Breast cancer, affecting both genders, but mostly females, exhibits shifting demographic patterns, with an increasing incidence in younger age groups. Early identification through mammography, clinical examinations, and breast self-exams enhances treatment efficacy, but challenges persist in low- and medium-income countries due to limited imaging resources. This review assesses the feasibility of employing breast ultrasound as the primary breast cancer screening method, particularly in resource-constrained regions. Following the PRISMA guidelines, this study examines 52 publications from the last five years. Breast ultrasound, distinct from mammography, offers advantages like radiation-free imaging, suitability for repeated screenings, and preference for younger populations. Real-time imaging and dense breast tissue evaluation enhance sensitivity, accessibility, and cost-effectiveness. However, limitations include reduced specificity, operator dependence, and challenges in detecting microcalcifications. Automatic breast ultrasound (ABUS) addresses some issues but faces constraints like potential inaccuracies and limited microcalcification detection. The analysis underscores the need for a comprehensive approach to breast cancer screening, emphasizing international collaboration and addressing limitations, especially in resource-constrained settings. Despite advancements, notably with ABUS, the primary goal is to contribute insights for optimizing breast cancer screening globally, improving outcomes, and mitigating the impact of this debilitating disease.

Limited access to breast cancer diagnosis globally leads to delayed treatment. Ultrasound, an effective yet underutilized method, requires specialized training for sonographers, which hinders its widespread use. Volume sweep imaging (VSI) is an innovative approach that enables untrained operators to capture high-quality ultrasound images. Combined with deep learning, like convolutional neural networks, it can potentially transform breast cancer diagnosis, enhancing accuracy, saving time and costs, and improving patient outcomes. The widely used UNet architecture, known for medical image segmentation, has limitations, such as vanishing gradients and a lack of multi-scale feature extraction and selective region attention. In this study, we present a novel segmentation model known as Wavelet&#95;Attention&#95;UNet (WATUNet). In this model, we incorporate wavelet gates and attention gates between the encoder and decoder instead of a simple connection to overcome the limitations mentioned, thereby improving model performance. Two datasets are utilized for the analysis: the public 'Breast Ultrasound Images' dataset of 780 images and a private VSI dataset of 3818 images, captured at the University of Rochester by the authors. Both datasets contained segmented lesions categorized into three types: no mass, benign mass, and malignant mass. Our segmentation results show superior performance compared to other deep networks. The proposed algorithm attained a Dice coefficient of 0.94 and an F1 score of 0.94 on the VSI dataset and scored 0.93 and 0.94 on the public dataset, respectively. Moreover, our model significantly outperformed other models in McNemar's test with false discovery rate correction on a 381-image VSI set. The experimental findings demonstrate that the proposed WATUNet model achieves precise segmentation of breast lesions in both standard-of-care and VSI images, surpassing state-of-the-art models. Hence, the model holds considerable promise for assisting in lesion identification, an essential step in the clinical diagnosis of breast lesions., Competing Interests: There is no conflict of interest., (© 2024 The Author(s). Published by IOP Publishing Ltd.)

Nonmass lesions in breast ultrasound (US) are areas of altered echogenicity without definite margins or mass effect. However, these lesions may show calcifications, associated architectural distortion, or shadowing just like masses. They vary in their echogenicity, distribution, ductal or nonductal appearance and the associated features that can be seen in variety of benign and malignant pathologies. With no uniform definition or classification system, there is no standardized approach in further risk categorization and management strategies of these lesions. Malignant nonmass lesions are not uncommon and few sonographic features can help in differentiating benign and malignant pathologies. US-guided tissue sampling or lesion localization can be preferred in the nonmass lesions identified on second look US after magnetic resonance imaging or mammography. This article aims to describe various imaging patterns and attempts to provide an algorithmic approach to nonmass findings on breast US., Competing Interests: Conflict of Interest None declared., (Indian Radiological Association. This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. ( https://creativecommons.org/licenses/by-nc-nd/4.0/ ).)

Abnormalities on breast ultrasound (US) images which do not meet the criteria for masses are referred to as nonmass lesions. These features and outcomes have been investigated in several studies conducted by Asian researchers. However, the term "nonmass" is not included in the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS) 5th edition for US. According to the Japan Association of Breast and Thyroid Sonology guidelines, breast lesions are divided into mass and nonmass. US findings of nonmass abnormalities are classified into five subtypes: abnormalities of the ducts, hypoechoic areas in the mammary glands, architectural distortion, multiple small cysts, and echogenic foci without a hypoechoic area. These findings can be benign or malignant; however, focal or segmental distributions and presence of calcifications suggest malignancy. Intraductal, invasive ductal, and lobular carcinomas can present as nonmass abnormalities. For the nonmass concept to be included in the next BI-RADS and be widely accepted in clinical practice, standardized terminologies, an interpretation algorithm, and outcome-based evidence are required for both screening and diagnostic US., Competing Interests: The authors have no potential conflicts of interest to disclose., (Copyright © 2024 The Korean Society of Radiology.)

We present subspecialty radiologist training for breast imaging at an Academic center in Dar Es Salaam, Tanzania. The training incorporates remote, in-person, asynchronous and synchronous teaching methods and multidisciplinary conferences. We use a team of US academic faculty under the auspices of the Radiological Society of North America Global Learning centers paradigm. Trainees are Tanzanian radiologists who are pursuing an additional specialization degree in Women's imaging, utilizing an approved 2-year curriculum. Challenges and opportunities in providing image-guided intervention and diagnosis training in the low- and middle-income settings are presented., (Copyright © 2023 Elsevier Inc. All rights reserved.)

Adenomyoepithelioma (AME) is a rare, usually benign breast neoplasm with low potential for malignant transformation. Imaging features are nonspecific and overlap with other benign and malignant breast lesions. On mammography, AME most often presents as a mass, usually oval in shape, with variable reported margins. Less commonly, AME can present mammographically as an asymmetry or can be mammographically occult. Associated calcifications are uncommon. On US, AME is usually seen as a hypoechoic oval mass, but it can also manifest as a complex cystic and solid mass. On US, the majority of AME have noncircumscribed margins (indistinct, angular, or microlobulated). Internal vascularity is usually present, and posterior enhancement can be seen. Although there is limited literature on MRI features, the most frequent finding is an irregular mass with washout kinetics; T2 hyperintensity can be observed. These nonspecific and often suspicious imaging features usually merit biopsy. On histologic analysis, AME is characterized by a biphasic proliferation of myoepithelial and epithelial cells. Pathologic diagnosis can be difficult due to the variety of histologic features of AME and heterogeneity in these tumors, especially when sampling is limited, such as in core needle biopsies. Wide local surgical excision of AME is recommended due to potential for recurrence and malignant transformation., (© Society of Breast Imaging 2023. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

The primary aim of our study was to assess the main mammographic and ultrasonographic features of invasive male breast malignancies. The secondary aim was to evaluate whether a specific radiological presentation would be associated with a worse receptor profile. Radiological images (mammography and/or ultrasound) of all patients who underwent surgery for male invasive breast cancer in our institution between 2008 and 2023 were retrospectively analyzed by two breast radiologists in consensus. All significant features of radiological presentation known in the literature were re-evaluated. Fifty-six patients were selected. The mean age at surgery of patients was 69 years (range: 35-81); in 82% of cases (46 patients), the histologic outcome was invasive ductal carcinoma. A total of 28 out of 56 (50%) patients had preoperative mammography; in 9/28 cases (32%), we found a mass with microcalcifications on mammography. The mass presented high density in 25 out of 28 patients (89%); the mass showed irregular margins in 15/28 (54%) cases. A total of 46 out of 56 patients had preoperative ultrasounds. The lesion showed a solid mass in 41/46 (89%) cases. In 5/46 patients (11%), the lesion was a mass with a mixed (partly liquid-partly solid) structure. We did not find any statistically significant correlation between major types of radiological presentation and tumor receptor arrangement. Knowledge of the main radiologic presentation patterns of malignant male breast neoplasm can help better manage this type of disease, which is rare but whose incidence is increasing.

Background: Breast ultrasound (US) is useful for dense breasts, and the introduction of artificial intelligence (AI)-assisted diagnoses of breast US images should be considered. However, the implementation of AI-based technologies in clinical practice is problematic because of the costs of introducing such approaches to hospital information systems (HISs) and the security risk of connecting HIS to the Internet to access AI services. To solve these problems, we developed a system that applies AI to the analysis of breast US images captured using a smartphone., Methods: Training data were prepared using 115 images of benign lesions and 201 images of malignant lesions acquired at the Division of Breast Surgery, Gifu University Hospital. YOLOv3 (object detection models) was used to detect lesions on US images. A graphical user interface (GUI) was developed to predict an AI server. A smartphone application was also developed for capturing US images displayed on the HIS monitor with its camera and displaying the prediction results received from the AI server. The sensitivity and specificity of the prediction performed on the AI server and via the smartphone were calculated using 60 images spared from the training., Results: The established AI showed 100% sensitivity and 75% specificity for malignant lesions and took 0.2 s per prediction with the AI sever. Prediction using a smartphone required 2 s per prediction and showed 100% sensitivity and 97.5% specificity for malignant lesions., Conclusions: Good-quality predictions were obtained using the AI server. Moreover, the quality of the prediction via the smartphone was slightly better than that on the AI server, which can be safely and inexpensively introduced into HISs., (© 2023. The Author(s).)

Objective: Aim: The objective of this literature review was to determine the optimal diagnostic algorithm for breast cancer detection associated with architectural distortion., Patients and Methods: Materials and Methods: The Pubmed, Google Scholar, Web of Science, and Scopus databases was used to search for materials on architectural distortion, associated pathologies, and radiological imaging methods., Conclusion: Conclusions: Architectural distortions may represent both benign and malignant pathology, but ultrasound, magnetic resonance, and roentgenological signs may not always with conf i dence determine the etiology of these changes. Therefore, the correct diagnostic algorithm which uses contrast imaging methods and digital breast tomosynthesis can prevent potentially exceeding radiation exposure, and unnecessary biopsies, change surgical operation volumes, decrease patient stress from procedures, and reduce the number of resources spent. After analyzing a number of scientif i c sources, we recommend morphological verif i -cation of all architectural distortions that are not the result of surgery, trauma, or that accumulate contrast media during breast imaging.

Rationale and Objectives: To develop a deep learning model for the automated classification of breast ultrasound images as benign or malignant. More specifically, the application of vision transformers, ensemble learning, and knowledge distillation is explored for breast ultrasound classification., Materials and Methods: Single view, B-mode ultrasound images were curated from the publicly available Breast Ultrasound Image (BUSI) dataset, which has categorical ground truth labels (benign vs malignant) assigned by radiologists and malignant cases confirmed by biopsy. The performance of vision transformers (ViT) is compared to convolutional neural networks (CNN), followed by a comparison between supervised, self-supervised, and randomly initialized ViT. Subsequently, the ensemble of 10 independently trained ViT, where the ensemble model is the unweighted average of the output of each individual model is compared to the performance of each ViT alone. Finally, we train a single ViT to emulate the ensembled ViT using knowledge distillation., Results: On this dataset that was trained using five-fold cross validation, ViT outperforms CNN, while self-supervised ViT outperform supervised and randomly initialized ViT. The ensemble model achieves an area under the receiver operating characteristics curve (AuROC) and area under the precision recall curve (AuPRC) of 0.977 and 0.965 on the test set, outperforming the average AuROC and AuPRC of the independently trained ViTs (0.958 ± 0.05 and 0.931 ± 0.016). The distilled ViT achieves an AuROC and AuPRC of 0.972 and 0.960., Conclusion: Both transfer learning and ensemble learning can each offer increased performance independently and can be sequentially combined to collectively improve the performance of the final model. Furthermore, a single vision transformer can be trained to match the performance of an ensemble of a set of vision transformers using knowledge distillation., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.)

Breast ultrasound (BUS) image classification in benign and malignant classes is often based on pre-trained convolutional neural networks (CNNs) to cope with small-sized training data. Nevertheless, BUS images are single-channel gray-level images, whereas pre-trained CNNs learned from color images with red, green, and blue (RGB) components. Thus, a gray-to-color conversion method is applied to fit the BUS image to the CNN's input layer size. This paper evaluates 13 gray-to-color conversion methods proposed in the literature that follow three strategies: replicating the gray-level image to all RGB channels, decomposing the image to enhance inherent information like the lesion's texture and morphology, and learning a matching layer. Besides, we introduce an image decomposition method based on the lesion's structural information to describe its inner and outer complexity. These gray-to-color conversion methods are evaluated under the same experimental framework using a pre-trained CNN architecture named ResNet-18 and a BUS dataset with more than 3000 images. In addition, the Matthews correlation coefficient (MCC), sensitivity (SEN), and specificity (SPE) measure the classification performance. The experimental results show that decomposition methods outperform replication and learning-based methods when using information from the lesion's binary mask (obtained from a segmentation method), reaching an MCC value greater than 0.70 and specificity up to 0.92, although the sensitivity is about 0.80. On the other hand, regarding the proposed method, the trade-off between sensitivity and specificity is better balanced, obtaining about 0.88 for both indices and an MCC of 0.73. This study contributes to the objective assessment of different gray-to-color conversion approaches in classifying breast lesions, revealing that mask-based decomposition methods improve classification performance. Besides, the proposed method based on structural information improves the sensitivity, obtaining more reliable classification results on malignant cases and potentially benefiting clinical practice., (© 2023. International Federation for Medical and Biological Engineering.)

Multi-task learning (MTL) methods have been extensively employed for joint localization and classification of breast lesions on ultrasound images to assist in cancer diagnosis and personalized treatment. One typical paradigm in MTL is a shared trunk network architecture. However, such a model design may suffer information-sharing conflicts and only achieve suboptimal performance for individual tasks. Additionally, the model relies on fully-supervised learning methodologies, imposing heavy burdens on data annotation. In this study, we propose a novel joint localization and classification model based on attention mechanisms and a sequential semi-supervised learning strategy to address these challenges. Our proposed framework offers three primary advantages. First, a lesion-aware network with multiple attention modules is designed to improve model performance on lesion localization. An attention-based classifier explicitly establishes correlations between the two tasks, alleviating information-sharing conflicts while leveraging location information to assist in classification. Second, a two-stage sequential semi-supervised learning strategy is designed for model training to achieve optimal performance on both tasks and substantially reduces the need for data annotation. Third, the asymmetric and modular model architecture allows for the flexible interchangeability of individual components, rendering the model adaptable to various applications. Experimental results from two different breast ultrasound image datasets under varied conditions have demonstrated the effectiveness of the proposed method. Furthermore, we conduct comprehensive investigations into the impacts of various factors on model performance, gaining in-depth insights into the mechanism of our proposed framework. The code is available at https://github.com/comp-imaging-sci/lanet-bus.git., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Breast tubular adenomas (TAs) are rare, benign glandular epithelial tumors that arise from a proliferation of acini in the terminal duct lobular units. In the literature, 40 TA cases have previously been reported, and we describe 5 additional cases in this article. In the small number of reported cases, TAs present most often in women of reproductive age but may also occur in postmenopausal women. Mammographically and sonographically, TAs are almost indistinguishable from fibroadenomas (FAs), and they typically present on US as hypoechoic, oval, circumscribed, parallel masses with variable internal vascularity. TAs can also be seen on mammography as oval masses with microlobulated margins, or as grouped coarse, heterogeneous microcalcifications with or without associated mass or asymmetry. On MRI, TAs present as heterogeneously enhancing, T2-hyperintense oval masses with persistent kinetics. Histopathologically, TAs consist of closely packed round tubules with minimal stroma, in distinction to FAs, which have a prominent stromal component that surrounds and can distort the associated tubules. Because of their benign classification and excellent prognosis, patients with biopsy-confirmed TAs may resume routine screening. Complete surgical excision may be considered for cosmetic purposes or for TAs exhibiting associated suspicious calcifications or rapid growth., (© Society of Breast Imaging 2023. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

It is important to make an appropriate diagnosis category by evaluating the morphology, blood flow, and hardness of mammary gland disease in breast ultrasound, such as screening and detailed examination. Therefore, understanding the characteristics and effects of B-mode (BM), color Doppler (CD), and elastography (EG) in ultrasonography is considered to lead to improvement in examination accuracy. In this study, we clarified the impact of each function on qualitative diagnosis, with the aim of assisting technologists with high-accuracy breast ultrasound. The subjects were 531 malignant tumor cases. The final evaluation of ultrasonography was classified into 5 levels, BM was classified into 5 levels, CD was classified into 4 levels based on blood flow evaluation, and EG was classified into 9 levels based on the Tsukuba elasticity score. The contribution rates of BM, CD, and EG were 70.4%, 15.0%, and 14.6% in all cases (correlation ratio 0.87). BM was most important at the time of breast ultrasound. No difference was observed between CD and EG. In the evaluation by malignant tumor, each characteristic was seen in the degree of influence. It became clear that comprehensive evaluation is important for benign/malignant evaluation and histological-type estimation.

The objective of this research was to study the contrast enhancement patterns of the different molecular subtypes of breast cancer on contrast-enhanced ultrasound (CEUS) using both qualitative and quantitative parameters. This prospective study included females with a single breast mass which was histopathologically proven carcinoma. B mode ultrasound (USG) and CEUS were performed in all patients during baseline assessment. Qualitative CEUS assessment encompassed enhancement pattern, presence of fill-in and washout. Quantitative assessment included measurement of peak enhancement, time to peak; area under the curve and mean transit time. A p -value < 0.05 was considered statistically significant for differentiating the subtypes. The included thirty masses were categorised into two subtypes-triple negative breast cancer (TNBC) (36.7%) and non-TNBC (63.3%) subtypes. With B-mode USG, a statistically significant difference was observed between the two groups with respect to their shape and margins. TNBC lesions showed an oval shape, circumscribed margins and peripheral nodular enhancement on CEUS with the absence of fill-in even in the delayed phase ( p -value - 0.04). The two subtypes did not significantly differ in terms of quantitative perfusion parameters. The various subtypes of breast cancer therefore possess distinct contrast enhancement patterns. CEUS potentially allows differentiation amongst these molecular subtypes that may aid in radiology-pathology (rad-path) correlation and follow up of the patients., Competing Interests: All authors declare that no conflicts exist., (© the authors; licensee ecancermedicalscience.)

Purpose: Breast ultrasound suffers from low positive predictive value and specificity. Artificial intelligence (AI) proposes to improve accuracy, reduce false negatives, reduce inter- and intra-observer variability and decrease the rate of benign biopsies. Perpetuating racial/ethnic disparities in healthcare and patient outcome is a potential risk when incorporating AI-based models into clinical practice; therefore, it is necessary to validate its non-bias before clinical use., Approach: Our retrospective review assesses whether our AI decision support (DS) system demonstrates racial/ethnic bias by evaluating its performance on 1810 biopsy proven cases from nine breast imaging facilities within our health system from January 1, 2018 to October 28, 2021. Patient age, gender, race/ethnicity, AI DS output, and pathology results were obtained., Results: Significant differences in breast pathology incidence were seen across different racial and ethnic groups. Stratified analysis showed that the difference in output by our AI DS system was due to underlying differences in pathology incidence for our specific cohort and did not demonstrate statistically significant bias in output among race/ethnic groups, suggesting similar effectiveness of our AI DS system among different races ( p > 0.05 for all)., Conclusions: Our study shows promise that an AI DS system may serve as a valuable second opinion in the detection of breast cancer on diagnostic ultrasound without significant racial or ethnic bias. AI tools are not meant to replace the radiologist, but rather to aid in screening and diagnosis without perpetuating racial/ethnic disparities., (© 2023 Society of Photo-Optical Instrumentation Engineers (SPIE).)

Deep learning (DL) models in breast ultrasound (BUS) image analysis face challenges with data imbalance and limited atypical tumor samples. Generative Adversarial Networks (GAN) address these challenges by providing efficient data augmentation for small datasets. However, current GAN approaches fail to capture the structural features of BUS and generated images lack structural legitimacy and are unrealistic. Furthermore, generated images require manual annotation for different downstream tasks before they can be used. Therefore, we propose a two-stage GAN framework, 2s-BUSGAN, for generating annotated BUS images. It consists of the Mask Generation Stage (MGS) and the Image Generation Stage (IGS), generating benign and malignant BUS images using corresponding tumor contours. Moreover, we employ a Feature-Matching Loss (FML) to enhance the quality of generated images and utilize a Differential Augmentation Module (DAM) to improve GAN performance on small datasets. We conduct experiments on two datasets, BUSI and Collected. Moreover, results indicate that the quality of generated images is improved compared with traditional GAN methods. Additionally, our generated images underwent evaluation by ultrasound experts, demonstrating the possibility of deceiving doctors. A comparative evaluation showed that our method also outperforms traditional GAN methods when applied to training segmentation and classification models. Our method achieved a classification accuracy of 69% and 85.7% on two datasets, respectively, which is about 3% and 2% higher than that of the traditional augmentation model. The segmentation model trained using the 2s-BUSGAN augmented datasets achieved DICE scores of 75% and 73% on the two datasets, respectively, which were higher than the traditional augmentation methods. Our research tackles imbalanced and limited BUS image data challenges. Our 2s-BUSGAN augmentation method holds potential for enhancing deep learning model performance in the field.