Digital mammography imaging: breast tomosynthesis and advanced applications

Digital Mammography, Tomosynthesis, and Contrast-Enhanced Mammography: Intraindividual Comparison of Mean Glandular Dose for Screening Examinations

BACKGROUND. Contrast-enhanced mammography (CEM) is growing in clinical use due to its increased sensitivity and specificity compared with full-field digital mammography (FFDM) and/or digital breast tomosynthesis (DBT), particularly in patients with dense breasts. OBJECTIVE. The purpose of this study was to perform an intraindividual comparison of mean glandular dose (MGD) with FFDM, DBT, a combination protocol using both FFDM and DBT (hereafter, combined FFDM-DBT), and CEM in patients undergoing breast cancer screening. METHODS. This retrospective study included 389 women (median age, 57.4 years) with an elevated risk of breast cancer who, as part of participation in an earlier prospective clinical trial, underwent breast cancer screening with combined FFDM-DBT and CEM between February 2019 and April 2021. A total of 764 breasts (383 left, 381 right) were evaluated. One craniocaudal (CC) view and one mediolateral oblique (MLO) view were evaluated per breast for each of FFDM, DBT, and CEM. MGD values were extracted from DICOM metadata. BI-RADS breast density categories were extracted from clinical radiology reports. Data were summarized descriptively, including determination of corresponding effective doses. RESULTS. The breast density category was A in zero patients, B in 44 patients (88 breasts), C in 306 patients (599 breasts), and D in 39 patients (77 breasts). The median MGD per breast (CC and MLO views combined) was 4.07 mGy for FFDM alone, 4.97 mGy for DBT alone, 9.38 mGy for combined FFDM-DBT, 3.96 mGy for low-energy CEM, 1.90 mGy for high-energy CEM, and 5.87 for CEM overall. Corresponding effective dose values were 0.49, 0.60, 1.13, 0.48, 0.23, and 0.70 mSv, respectively. The median MGD for density categories B, C, and D, respectively, was 4.01, 4.22, and 2.70 mGy for FFDM; 5.93, 4.93, and 3.17 mGy for DBT; and 5.90, 6.02, and 4.52 mGy for CEM. CONCLUSION. In this intraindividual comparative study of screening examinations, the MGD per breast was higher for CEM than for FFDM or DBT alone. However, these differences were small, and MGD was lower for CEM than for combined FFDM-DBT. CLINICAL IMPACT. These findings are relevant to ongoing considerations of the role of CEM in breast cancer screening pathways.

Cross-Attention Adaptive Feature Pyramid Network with Uncertainty Boundary Modeling for Mass Detection in Digital Breast Tomosynthesis

Computer-aided detection (CADe) of masses in digital breast tomosynthesis (DBT) is crucial for early breast cancer diagnosis. However, the variability in the size and morphology of breast masses and their resemblance to surrounding tissues present significant challenges. Current CNN-based CADe methods, particularly those that use Feature Pyramid Networks (FPN), often fail to integrate multi-scale information effectively and struggle to handle dense glandular tissue with high-density or iso-density mass lesions due to the unidirectional integration and progressive attenuation of features, leading to high false positive rates. Additionally, the commonly indistinct boundaries of breast masses introduce uncertainty in boundary localization, which makes traditional Dirac boundary modeling insufficient for precise boundary regression. To address these issues, we propose the CU-Net network, which efficiently fuses multi-scale features and accurately models blurred boundaries. Specifically, the CU-Net introduces the Cross-Attention Adaptive Feature Pyramid Network (CA-FPN), which enhances the effectiveness and accuracy of feature interactions through a cross-attention mechanism to capture global correlations across multi-scale feature maps. Simultaneously, the Breast Density Perceptual Module (BDPM) incorporates breast density information to weight intermediate features, thereby improving the network's focus on dense breast regions susceptible to false positives. For blurred mass boundaries, we introduce Uncertainty Boundary Modeling (UBM) to model the positional distribution function of predicted bounding boxes for masses with uncertain boundaries. In comparative experiments on an in-house clinical DBT dataset and the BCS-DBT dataset, the proposed method achieved sensitivities of 89.68% and 72.73% at 2 false positives per DBT volume (FPs/DBT), respectively, significantly outperforming existing state-of-the-art detection methods. This method offers clinicians rapid, accurate, and objective diagnostic assistance, demonstrating substantial potential for clinical application.

Greater benefits of deep learning-based computer-aided detection systems for finding small signals in 3D volumetric medical images

Purpose

Radiologists are tasked with visually scrutinizing large amounts of data produced by 3D volumetric imaging modalities. Small signals can go unnoticed during the 3D search because they are hard to detect in the visual periphery. Recent advances in machine learning and computer vision have led to effective computer-aided detection (CADe) support systems with the potential to mitigate perceptual errors.

Approach

Sixteen nonexpert observers searched through digital breast tomosynthesis (DBT) phantoms and single cross-sectional slices of the DBT phantoms. The 3D/2D searches occurred with and without a convolutional neural network (CNN)-based CADe support system. The model provided observers with bounding boxes superimposed on the image stimuli while they looked for a small microcalcification signal and a large mass signal. Eye gaze positions were recorded and correlated with changes in the area under the ROC curve (AUC).

Results

The CNN-CADe improved the 3D search for the small microcalcification signal ( Δ   AUC = 0.098 , p = 0.0002 ) and the 2D search for the large mass signal ( Δ   AUC = 0.076 , p = 0.002 ). The CNN-CADe benefit in 3D for the small signal was markedly greater than in 2D ( Δ Δ   AUC = 0.066 , p = 0.035 ). Analysis of individual differences suggests that those who explored the least with eye movements benefited the most from the CNN-CADe ( r = - 0.528 , p = 0.036 ). However, for the large signal, the 2D benefit was not significantly greater than the 3D benefit ( Δ Δ   AUC = 0.033 , p = 0.133 ).

Conclusion

The CNN-CADe brings unique performance benefits to the 3D (versus 2D) search of small signals by reducing errors caused by the underexploration of the volumetric data.

Impact of Imaging Biomarkers and AI on Breast Cancer Management: A Brief Review

Breast cancer stands out as the most frequently identified malignancy, ranking as the fifth leading cause of global cancer-related deaths. The American College of Radiology (ACR) introduced the Breast Imaging Reporting and Data System (BI-RADS) as a standard terminology facilitating communication between radiologists and clinicians; however, an update is now imperative to encompass the latest imaging modalities developed subsequent to the 5th edition of BI-RADS. Within this review article, we provide a concise history of BI-RADS, delve into advanced mammography techniques, ultrasonography (US), magnetic resonance imaging (MRI), PET/CT images, and microwave breast imaging, and subsequently furnish comprehensive, updated insights into Molecular Breast Imaging (MBI), diagnostic imaging biomarkers, and the assessment of treatment responses. This endeavor aims to enhance radiologists' proficiency in catering to the personalized needs of breast cancer patients. Lastly, we explore the augmented benefits of artificial intelligence (AI), machine learning (ML), and deep learning (DL) applications in segmenting, detecting, and diagnosing breast cancer, as well as the early prediction of the response of tumors to neoadjuvant chemotherapy (NAC). By assimilating state-of-the-art computer algorithms capable of deciphering intricate imaging data and aiding radiologists in rendering precise and effective diagnoses, AI has profoundly revolutionized the landscape of breast cancer radiology. Its vast potential holds the promise of bolstering radiologists' capabilities and ameliorating patient outcomes in the realm of breast cancer management.

The Impact of Adding Digital Breast Tomosynthesis to BI-RADS Categorization of Mammographically Equivocal Breast Lesions

Digital mammography (DM) is the cornerstone of breast cancer detection. Digital breast tomosynthesis (DBT) is an advanced imaging technique used for diagnosing and screening breast lesions, particularly in dense breasts. This study aimed to evaluate the impact of combining DBT with DM on the BI-RADS categorization of equivocal breast lesions. We prospectively evaluated 148 females with equivocal BI-RADS breast lesions (BI-RADS 0, 3, and 4) with DM. All patients underwent DBT. Two experienced radiologists analyzed the lesions. They then assigned a BI-RADS category for each lesion according to the BI-RADS 2013 lexicon, using DM, DBT, and integrated DM and DBT. We compared the results based on major radiological characteristics, BI-RADS classification, and diagnostic accuracy, using the histopathological examination of the lesions as a reference standard. The total number of lesions was 178 on DBT and 159 on DM. Nineteen lesions were discovered using DBT and were missed by DM. The final diagnoses of 178 lesions were malignant (41.6%) and benign (58.4%). Compared to DM, DBT produced 34.8% downgrading and 32% upgrading of breast lesions. Compared with DM, DBT decreased the number of BI-RADS 4 and 3. All the upgraded BI-RADS 4 lesions were confirmed to be malignant. The combination of DM and DBT improves the diagnostic accuracy of BI-RADS for evaluating and characterizing mammographic equivocal breast lesions and allows for proper BI-RADS categorization.

Artificial intelligence assistance for women who had spot compression view: reducing recall rates for digital mammography

Background

Mammography yields inevitable recall for indeterminate findings that need to be confirmed with additional views.

Purpose

To explore whether the artificial intelligence (AI) algorithm for mammography can reduce false-positive recall in patients who undergo the spot compression view.

Material and Methods

From January to December 2017, 236 breasts from 225 women who underwent the spot compression view due to focal asymmetry, mass, or architectural distortion on standard digital mammography were included. Three readers who were blinded to the study purpose, patient information, previous mammograms, following spot compression views, and any clinical or pathologic reports retrospectively reviewed 236 standard mammograms and determined the necessity of patient recall and the probability of malignancy per breast, first without and then with AI assistance. The performances of AI and the readers were evaluated with the recall rate, area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy.

Results

Among 236 examinations, 8 (3.4%) were cancers and 228 (96.6%) were benign. The recall rates of all three readers significantly decreased with AI assistance ( P < 0.05). The reader-averaged recall rates significantly decreased with AI assistance regardless of breast composition (fatty breasts: 32.7% to 24.1%m P = 0.002; dense breasts: 33.6% to 21.2%, P < 0.001). The reader-averaged AUC increased with AI assistance and was comparable to that of standalone AI (0.835 vs. 0.895; P = 0.234). The reader-averaged specificity (71.2% to 79.8%, P < 0.001) and accuracy (71.3% to 79.7%, P  < 0.001) significantly improved with AI assistance.

Conclusion

AI assistance significantly reduced false-positive recall without compromising cancer detection in women with focal asymmetry, mass, or architectural distortion on standard digital mammography regardless of mammographic breast density.

Current and future diagnostic and treatment strategies for patients with invasive lobular breast cancer

BACKGROUND

Invasive lobular breast cancer (ILC) is the second most common type of breast cancer after invasive breast cancer of no special type (NST), representing up to 15% of all breast cancers.

DESIGN

Latest data on ILC are presented, focusing on diagnosis, molecular make-up according to the European Society for Medical Oncology Scale for Clinical Actionability of molecular Targets (ESCAT) guidelines, treatment in the early and metastatic setting and ILC-focused clinical trials.

RESULTS

At the imaging level, magnetic resonance imaging-based and novel positron emission tomography/computed tomography-based techniques can overcome the limitations of currently used imaging techniques for diagnosing ILC. At the pathology level, E-cadherin immunohistochemistry could help improving inter-pathologist agreement. The majority of patients with ILC do not seem to benefit as much from (neo-)adjuvant chemotherapy as patients with NST, although chemotherapy might be required in a subset of high-risk patients. No differences in treatment efficacy are seen for anti-human epidermal growth factor receptor 2 (HER2) therapies in the adjuvant setting and cyclin-dependent kinases 4 and 6 inhibitors in the metastatic setting. The clinical utility of the commercially available prognostic gene expression-based tests is unclear for patients with ILC. Several ESCAT alterations differ in frequency between ILC and NST. Germline BRCA1 and PALB2 alterations are less frequent in patients with ILC, while germline CDH1 (gene coding for E-cadherin) alterations are more frequent in patients with ILC. Somatic HER2 mutations are more frequent in ILC, especially in metastases (15% ILC versus 5% NST). A high tumour mutational burden, relevant for immune checkpoint inhibition, is more frequent in ILC metastases (16%) than in NST metastases (5%). Tumours with somatic inactivating CDH1 mutations may be vulnerable for treatment with ROS1 inhibitors, a concept currently investigated in early and metastatic ILC.

CONCLUSION

ILC is a unique malignancy based on its pathological and biological features leading to differences in diagnosis as well as in treatment response, resistance and targets as compared to NST.

Intelligent Computer-Aided Model for Efficient Diagnosis of Digital Breast Tomosynthesis 3D Imaging Using Deep Learning

Digital breast tomosynthesis (DBT) is a highly promising 3D imaging modality for breast diagnosis. Tissue overlapping is a challenge with traditional 2D mammograms; however, since digital breast tomosynthesis can obtain three-dimensional images, tissue overlapping is reduced, making it easier for radiologists to detect abnormalities and resulting in improved and more accurate diagnosis. In this study, a new computer-aided multi-class diagnosis system is proposed that integrates DBT augmentation and colour feature map with a modified deep learning architecture (Mod_AlexNet). To the proposed modified deep learning architecture (Mod AlexNet), an optimization layer with multiple high performing optimizers is incorporated so that it can be evaluated and optimised using various optimization techniques. Two experimental scenarios are applied, the first scenario proposed a computer-aided diagnosis (CAD) model that integrated DBT augmentation, image enhancement techniques and colour feature mapping with six deep learning models for feature extraction, including ResNet-18, AlexNet, GoogleNet, MobileNetV2, VGG-16 and DenseNet-201, to efficiently classify DBT slices. The second scenario compared the performance of the newly proposed Mod_AlexNet architecture and traditional AlexNet, using several optimization techniques and different evaluation performance metrics were computed. The optimization techniques included adaptive moment estimation (Adam), root mean squared propagation (RMSProp), and stochastic gradient descent with momentum (SGDM), for different batch sizes, including 32, 64 and 512. Experiments have been conducted on a large benchmark dataset of breast tomography scans. The performance of the first scenario was compared in terms of accuracy, precision, sensitivity, specificity, runtime, and f1-score. While in the second scenario, performance was compared in terms of training accuracy, training loss, and test accuracy. In the first scenario, results demonstrated that AlexNet reported improvement rates of 1.69%, 5.13%, 6.13%, 4.79% and 1.6%, compared to ResNet-18, MobileNetV2, GoogleNet, DenseNet-201 and VGG16, respectively. Experimental analysis with different optimization techniques and batch sizes demonstrated that the proposed Mod_AlexNet architecture outperformed AlexNet in terms of test accuracy with improvement rates of 3.23%, 1.79% and 1.34% when compared using SGDM, Adam, and RMSProp optimizers, respectively.

Advances in breast cancer screening modalities and status of global screening programs

Breast cancer (BC) is the most prevalent malignancy worldwide, and a continued upward trend has been predicted in the coming decades. Screening in selected targeted populations, which is effective in reducing cancer-related mortality, has been widely implemented in many countries. This review summarizes the advances in BC screening techniques, organized or opportunistic BC screening programs across different countries, and screening modalities recommended by different academic authorities. Mammography is the most widely used and effective technique for BC screening. Other complementary techniques include ultrasound, clinical breast examination, and magnetic resonance imaging. Novel screening tests, including digital breast tomosynthesis and liquid biopsies, are still under development. Globally, the implementation status of BC screening programs is uneven, which is reflected by differences in screening modes, techniques, and population coverage. The recommended optimal screening strategies varied according to the authoritative guidelines. The effectiveness of current screening programs is influenced by several factors, including low detection rate, high false-positive rate, and unsatisfactory coverage and uptake rates. Exploration of accurate BC risk prediction models and the development of risk-stratified screening strategies are highly warranted in future research.

Variation in digital breast tomosynthesis image quality at differing heights above the detector

INTRODUCTION

The aim of this preliminary work was to determine if image quality in digital breast tomosynthesis (DBT) changes when tomosynthesis image slices were obtained at differing heights above the detector and in differing breast thicknesses.

METHODS

A CIRS Model 020 BR3D breast imaging phantom was used to obtain the DBT images. The images were also acquired at different tube voltages, and each exposure was determined by the automatic exposure control system. Contrast-to-noise ratio (CNR) and figure-of-merit (FOM) values were obtained and compared.

RESULTS

At a phantom thickness of 5 cm or greater, there was a significant reduction (P ≤ 0.05) of image CNR values obtained from the images near the top of the phantom to those obtained near the bottom of the phantom. When the phantom thickness was 4 cm, there was no significant difference in CNR values between DBT images acquired at any height in the phantom. FOM values generally showed no difference when images were obtained at differing heights above the detector.

CONCLUSION

Image quality, as measured by the CNR, was reduced when tomosynthesis slice image heights were at the top of the phantom and when the thickness of the phantom was more than 4 cm. From this preliminary work, clinicians need to be aware that DBT images obtained near the top of the breast, when breast thickness is greater than 4 cm, may have reduced image quality. Further work is needed to fully assess any DBT image quality changes when images are obtained near the top of the breast.

Diagnostic Challenge of Invasive Lobular Carcinoma of the Breast: What Is the News? Breast Magnetic Resonance Imaging and Emerging Role of Contrast-Enhanced Spectral Mammography

Invasive lobular carcinoma is the second most common histologic form of breast cancer, representing 5% to 15% of all invasive breast cancers. Due to an insidious proliferative pattern, invasive lobular carcinoma remains clinically and radiologically elusive in many cases. Breast magnetic resonance imaging (MR) is considered the most accurate imaging modality in detecting and staging invasive lobular carcinoma and it is strongly recommended in pre-operative planning for all ILC. Contrast-enhanced spectral mammography (CESM) is a new diagnostic method that enables the accurate detection of malignant breast lesions similar to that of breast MR. CESM is also a promising breast imaging method for planning surgeries. In this study, we compare the ability of contrast-enhanced spectral mammography (CESM) with breast MR in the preoperative assessment of the extent of invasive lobular carcinoma. All patients with proven invasive lobular carcinoma treated in our breast cancer center underwent preoperative breast MRI and CESM. Images were reviewed by two dedicated breast radiologists and results were compared to the reference standard histopathology. CESM was similar and in some cases more accurate than breast MR in assessing the extent of disease in invasive lobular cancers. Further evaluation in larger prospective randomized trials is needed to validate our preliminary results.

A Realistic Breast Phantom Proposal for 3D Image Reconstruction in Digital Breast Tomosynthesis

Objectives: Iterative (eg, simultaneous algebraic reconstruction technique [SART]) and analytical (eg, filtered back projection [FBP]) image reconstruction techniques have been suggested to provide adequate three-dimensional (3D) images of the breast for capturing microcalcifications in digital breast tomosynthesis (DBT). To decide on the reconstruction method in clinical DBT, it must first be tested in a simulation resembling the real clinical environment. The purpose of this study is to introduce a 3D realistic breast phantom for determining the reconstruction method in clinical applications. Methods: We designed a 3D realistic breast phantom with varying dimensions (64³-512³) mimicking some structures of a real breast such as milk ducts, lobules, and ribs using TomoPhantom software. We generated microcalcifications, which mimic cancerous cells, with a separate MATLAB code and embedded them into the phantom for testing and benchmark studies in DBT. To validate the characterization of the phantom, we tested the distinguishability of microcalcifications by performing 3D image reconstruction methods (SART and FBP) using Laboratory of Computer Vision (LAVI) open-source reconstruction toolbox. Results: The creation times of the proposed realistic breast phantom were seconds of 2.5916, 8.4626, 57.6858, and 472.1734 for 64³, 128³, 256³, and 512³, respectively. We presented reconstructed images and quantitative results of the phantom for SART (1-2-4-8 iterations) and FBP, with 11 to 23 projections. We determined qualitatively and quantitatively that SART (2-4 iter.) yields better results than FBP. For example, for 23 projections, the contrast-to-noise ratio (CNR) values of SART (2 iter.) and FBP were 2.871 and 0.497, respectively. Conclusions: We created a computationally efficient realistic breast phantom that is eligible for reconstruction and includes anatomical structures and microcalcifications, successfully. By proposing this breast phantom, we provided the opportunity to test which reconstruction methods can be used in clinical applications vary according to various parameters such as the No. of iterations and projections in DBT.

Do automated breast ultrasound and tomosynthesis have an effective role in dense breast evaluation?

Background

Mammography plays a great role in reducing breast cancer mortality as it is the standard method of breast imaging and screening. But the accuracy of mammography performance reduces in cancer detection in women with dense breast due to the summation of images and overlapping of breast tissue. ABUS and tomosynthesis both recently help to detect breast cancer in dense breasted women. This prospective study was done in the female imaging unit and approved by its research and ethical committee; all the patients did an informed consent during the period from October 2018 to March 2019. The study was conducted on 38 patients with 38 lesions subjected to digital mammography, tomosynthesis and automated breast ultrasound (ABUS), who all had dense breast in mammography.

Results

Automated breast ultrasound (ABUS) showed 100% in all sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV) as well as accuracy, while the digital mammography tomosynthesis showed 100% in specificity, 87.5% in sensitivity, 100% in PPV, 82.4% in NPV and 92.1% accuracy.

Conclusion

Automated breast ultrasound (ABUS) together with tomosynthesis makes a revolution in breast screening and detecting cancer in women with dense breasts.

SCREENING MAMMOGRAPHY: DIAGNOSTIC EFFICACY-ISSUES AND CONSIDERATIONS FOR THE 2020S

Diagnostic efficacy in medical imaging is ultimately a reflection of radiologist performance. This can be influenced by numerous factors, some of which are patient related, such as the physical size and density of the breast, and machine related, where some lesions are difficult to visualise on traditional imaging techniques. Other factors are human reader errors that occur during the diagnostic process, which relate to reader experience and their perceptual and cognitive oversights. Given the large-scale nature of breast cancer screening, even small increases in diagnostic performance equate to large numbers of women saved. It is important to identify the causes of diagnostic errors and how detection efficacy can be improved. This narrative review will therefore explore the various factors that influence mammographic performance and the potential solutions used in an attempt to ameliorate the errors made.

Automated Classification of Breast Cancer Lesions for Digitised Mammograms via Computer-Aided Diagnosis System

Women with breast cancer have a high risk of death. Digitised mammograms can be used to detect the early stage of breast cancer. However, digitised mammograms suffer low contrast appearances that may lead to misdiagnosis. This paper proposes a Computer-Aided Diagnosis (CAD) system of automated classification of breast cancer lesions using a modified image processing technique of Fuzzy Anisotropic Diffusion Histogram Equalization Contrast Adaptive Limited (FADHECAL) incorporated with Multilevel Otsu Thresholding on digitised mammograms. Four main blocks were used in this CAD system, namely; (i) Pre-processing and Enhancement block; (ii) Segmentation block; (iii) Region of Interests (ROIs) Extraction block; and (iv) Classification block. The CAD system was tested on 30 digitised mammograms retrieved from the Mini-Mammographic Image Analysis Society (MIAS) database with various degrees of severity and background tissues. The proposed CAD system showed a high accuracy of 96.67% for the detection of breast cancer lesions.

Comparing the diagnostic efficacy of full field digital mammography with digital breast tomosynthesis using BIRADS score in a tertiary cancer care hospital

Introduction

Breast cancer is one of the leading cancers in females worldwide, and its incidence has been rising at an exponential pace in the last 10 years even in India. Mammography has been the mainstay for detection of breast cancer over decades and has gradually advanced from screen film to full-field digital mammography. Recently, tomosynthesis has evolved as an advanced imaging investigation for early diagnosis of breast lesions in both diagnostic and screening settings.

Aim of Study

To compare and evaluate the impact of digital breast tomosynthesis (DBT) compared to full-field digital mammography (FFDM) in the interpretation of BIRADS score in both diagnostic and screening settings.

Settings and Design

A 1-year prospective longitudinal study was conducted in the Department of Radio-diagnosis in our institute using Hologic Selenia Dimensions for mammography as well as tomosynthesis.

Materials and Methods

One hundred women known or suspected (opportunistic screening) for breast cancer were evaluated either with FFDM alone or both FFDM and DBT. Sensitivity, specificity, positive predictive value, negative predictive value, and P value were used to assess the various diagnostic criteria in our study.

Results

Addition of DBT to FFDM results in a statistically significant increase in the sensitivity, specificity, and positive predictive value, and a statistically significant decrease in the false positive rates. Similar results were noted in both diagnostic and screening cases. It was observed that, in most cases, i.e. a total of 47, DBT did not change the BIRADS scoring; however, its addition increased the diagnostic confidence. BIRADS was upgraded and downgraded in 14 and 31 cases, respectively, with the addition of DBT to FFDM. New lesions were seen with addition of DBT to FFDM in 8 cases.

Conclusion

Addition of DBT to FFDM results in increase in sensitivity, specificity, positive predictive value, and a statistically significant decrease in false positive rates in both diagnostic and screening cases. As addition of tomosynthesis results in a significant decrease in recall rate, it should be added, at least, in all screening mammography programs.

Impact of switching from digital mammography to tomosynthesis plus digital mammography on breast cancer screening in Alberta, Canada

OBJECTIVES

To compare abnormal call rates (ACR), cancer detection rates (CDR), positive predictive values (PPVs), and annual return to screen recommendations after switching from digital mammography (DM) to digital breast tomosynthesis plus DM (DBT + DM) for breast cancer screening.

SETTING

The Alberta Breast Cancer Screening Program collects screening data from clinics throughout the province of Alberta, Canada.

METHODS

This study retrospectively collected data, between 2015 and 2018, on women aged 40+ who underwent breast cancer screening at two large volume multisite radiology groups to compare metrics one year prior and one year after DBT + DM implementation. Comparisons between modalities were carried out within age groups, within breast density categories, and for initial vs. subsequent screens.

RESULTS

A total of 125,432 DM and 128,912 DBT + DM screening exams were performed. For women aged 50-74, the DBT + DM group had a higher ACR (p < 0.01) but lower annual return to screens (p < 0.01). CDR was higher post-DBT + DM implementation for women with scattered (6.0 per 1000 vs. 4.4 per 1000; p = 0.001) or heterogeneously dense breasts (6.5 per 1000 vs. 4.2 per 1000; p < 0.001). PPV was higher with DBT + DM for all age groups, with women 50-74 having a PPV of 8.3% using DBT + DM vs. 7.1% with DM (p = 0.009).

CONCLUSION

All metrics improved or stayed the same after switching to DBT + DM except for ACR. However, the increase in ACR could be attributed to a trend already occurring prior to the switch. Longer term monitoring is needed to confirm these findings.

Adherence to Screening Among American Indian Women Accessing a Mobile Mammography Unit

RATIONALE AND OBJECTIVES

Although screening mammography is essential to reducing breast cancer morbidity and mortality, barriers exist especially among underrepresented minority groups. There are few studies of mammogram screening among American Indian women, many of whom reside in rural areas where screening access is challenging. A mobile mammography unit served 24 Indian Health Service clinics during 2013-17. Screening mammography adherence was evaluated.

MATERIALS AND METHODS

Among mobile unit women, 'adherence to screening' was determined by the date of the most recent prior mammogram. Those having a prior mammogram 9-27 months ago were classified as 'adherent to screening'. Comparison screening data were obtained from the American College of Radiology National Mammography Database, consisting of screening cases occurring in year 2015. Additionally, among mobile unit women 'continued adherence to screening' was determined, defined as at least one repeat screening at the mobile unit within the subsequent 9-27 months after a screening there.

RESULTS

Among 1,615 mobile unit women, 624 (38.6%) were adherent to screening. Among 2,509,826 National Mammography Database women, 1,481,021 (59.0%) were adherent to screening. (p<0.0001) The prevalence of a >27-month interval between mammograms was 3.13 (95% CI 2.91-3.36) times greater among mobile unit women than National Mammography Database women. 'Continued adherence to screening' of mobile unit women was 428/1194 (35.9%).

CONCLUSION

Adherence to screening and continued adherence to screening were low among mobile unit women and time interval between screenings was longer than National Mammography Database women. Factors to improve screening adherence among these underserved women should be determined.

Pros and cons for breast cancer screening with tomosynthesis - a review of the literature

Breast cancer screening programs using mammography proved their value in detecting breast cancer at early stages and, consequently, reducing the mortality from this disease.

Due to the technological progress, the screening programs have shifted from screen-film mammography to digital mammography and nowadays digital breast tomosynthesis became the focus of breast imaging research. Using tomosynthesis in screening increases cancer detection rates and decreases recall and false-positive rates, thus improving the effectiveness of breast cancer screening programs, with positive consequences on health care costs and on patient psychology.

More long-term follow-up data must be collected for assessing absolute sensitivity and specificity of digital breast tomosynthesis, together with efforts for addressing the limitations of the method.

Breast cancer, screening and diagnostic tools: All you need to know

Breast cancer is one of the most frequent malignancies among women worldwide. Methods for screening and diagnosis allow health care professionals to provide personalized treatments that improve the outcome and survival. Scientists and physicians are working side-by-side to develop evidence-based guidelines and equipment to detect cancer earlier. However, the lack of comprehensive interdisciplinary information and understanding between biomedical, medical, and technology professionals makes innovation of new screening and diagnosis tools difficult. This critical review gathers, for the first time, information concerning normal breast and cancer biology, established and emerging methods for screening and diagnosis, staging and grading, molecular and genetic biomarkers. Our purpose is to address key interdisciplinary information about these methods for physicians and scientists. Only the multidisciplinary interaction and communication between scientists, health care professionals, technical experts and patients will lead to the development of better detection tools and methods for an improved screening and early diagnosis.

Imaging of Breast Cancers With Predilection for Nonmass Pattern of Growth: Invasive Lobular Carcinoma and DCIS-Does Imaging Capture It All?

OBJECTIVE. Invasive lobular carcinoma (ILC) and ductal carcinoma in situ (DCIS) are distinct histopathologic entities with several commonalities: both have subtle clinical and imaging presentation, have been linked with controversy regarding optimal imaging techniques and management, and exemplify the codependence of adequate imaging evaluation and optimal treatment strategies in breast care. CONCLUSION. We review molecular mechanisms and histopathologic patterns that define the biologic behavior of both ILC and DCIS and discuss how these mechanisms translate into distinct clinical and imaging presentations that affect the staging workup and patient management algorithm.

Long-Term Outcomes and Cost-Effectiveness of Breast Cancer Screening With Digital Breast Tomosynthesis in the United States

BACKGROUND

Digital breast tomosynthesis (DBT) is increasingly being used for routine breast cancer screening. We projected the long-term impact and cost-effectiveness of DBT compared to conventional digital mammography (DM) for breast cancer screening in the United States.

METHODS

Three Cancer Intervention and Surveillance Modeling Network breast cancer models simulated US women ages 40 years and older undergoing breast cancer screening with either DBT or DM starting in 2011 and continuing for the lifetime of the cohort. Screening performance estimates were based on observational data; in an alternative scenario, we assumed 4% higher sensitivity for DBT. Analyses used federal payer perspective; costs and utilities were discounted at 3% annually. Outcomes included breast cancer deaths, quality-adjusted life-years (QALYs), false-positive examinations, costs, and incremental cost-effectiveness ratios (ICERs).

RESULTS

Compared to DM, DBT screening resulted in a slight reduction in breast cancer deaths (range across models 0-0.21 per 1000 women), small increase in QALYs (1.97-3.27 per 1000 women), and a 24-28% reduction in false-positive exams (237-268 per 1000 women) relative to DM. ICERs ranged from $195 026 to $270 135 per QALY for DBT relative to DM. When assuming 4% higher DBT sensitivity, ICERs decreased to $130 533-$156 624 per QALY. ICERs were sensitive to DBT costs, decreasing to $78 731 to $168 883 and $52 918 to $118 048 when the additional cost of DBT was reduced to $36 and $26 (from baseline of $56), respectively.

CONCLUSION

DBT reduces false-positive exams while achieving similar or slightly improved health benefits. At current reimbursement rates, the additional costs of DBT screening are likely high relative to the benefits gained; however, DBT could be cost-effective at lower screening costs.

The added value of digital breast tomosynthesis in improving diagnostic performance of BI-RADS categorization of mammographically indeterminate breast lesions

Background

Mammographic findings are seen more clearly in tomographic images with consequent improvement of Breast Imaging Reporting and Data System (BI-RADS) in categorization of indeterminate breast lesions. This study aimed to evaluate the added value of digital breast tomosynthesis (DBT) to BI-RADS classification in categorization of indeterminate breast lesions after digital mammography (DM) as an initial approach.

Methods and results

We prospectively evaluated 296 women with BI-RADS indeterminate breast lesions (BI-RADS 0, 3, and 4) by DM between January 2018 and October 2019. All patients underwent DBT. Two radiologists evaluated lesions and assigned a BI-RADS category to each lesion according to BI-RADS lexicon 2013 classification using DM, DBT, and combined DM and DBT. The results were compared in terms of main radiological features, diagnostic performance, and BI-RADS classification using histopathology as the reference standard. A total of 355 lesions were detected on DBT and 318 lesions on DM. Thirty-seven lesions were detected by DBT and not seen by DM. The final diagnoses of 355 lesions were 58.3% benign and 41.7% malignant. In comparison to DM, DBT produced 31.5% upgrading and 35.2% downgrading of BI-RADS scoring of breast lesions. DBT reduced number of BI-RADS 3 and 4, compared to DM. All upgraded BI-RADS 4 were malignant. The combination of DBT and DM significantly increased the performance of BI-RADS in the diagnosis of indeterminate breast lesions versus DM or DBT alone (p < 0.001).

Conclusion

Adding DBT to BI-RADS improves its diagnostic performance in detection and characterization of mammography indeterminate breast lesions.

Breast Cancer Detection-A Synopsis of Conventional Modalities and the Potential Role of Microwave Imaging

Global statistics have demonstrated that breast cancer is the most frequently diagnosed invasive cancer and the leading cause of cancer death among female patients. Survival following a diagnosis of breast cancer is grossly determined by the stage of the disease at the time of initial diagnosis, highlighting the importance of early detection. Improving early diagnosis will require a multi-faceted approach to optimizing the use of currently available imaging modalities and investigating new methods of detection. The application of microwave technologies in medical diagnostics is an emerging field of research, with breast cancer detection seeing the most significant progress in the last twenty years. In this review, the application of current conventional imaging modalities is discussed, and recurrent shortcomings highlighted. Microwave imaging is rapid and inexpensive. If the preliminary results of its diagnostic capacity are substantiated, microwave technology may offer a non-ionizing, non-invasive, and painless adjunct or stand-alone modality that could possibly be implemented in routine diagnostic breast care.

DBT Masses Automatic Segmentation Using U-Net Neural Networks

To improve the automatic segmentation accuracy of breast masses in digital breast tomosynthesis (DBT) images, we propose a DBT mass automatic segmentation algorithm by using a U-Net architecture. Firstly, to suppress the background tissue noise and enhance the contrast of the mass candidate regions, after the top-hat transform of DBT images, a constraint matrix is constructed and multiplied with the DBT image. Secondly, an efficient U-Net neural network is built and image patches are extracted before data augmentation to establish the training dataset to train the U-Net model. And then the presegmentation of the DBT tumors is implemented, which initially classifies per pixel into two different types of labels. Finally, all regions smaller than 50 voxels considered as false positives are removed, and the median filter smoothes the mass boundaries to obtain the final segmentation results. The proposed method can effectively improve the performance in the automatic segmentation of the masses in DBT images. Using the detection Accuracy (Acc), Sensitivity (Sen), Specificity (Spe), and area under the curve (AUC) as evaluation indexes, the Acc, Sen, Spe, and AUC for DBT mass segmentation in the entire experimental dataset is 0.871, 0.869, 0.882, and 0.859, respectively. Our proposed U-Net-based DBT mass automatic segmentation system obtains promising results, which is superior to some classical architectures, and may be expected to have clinical application prospects.

Hybrid Optimization Method (HOM) Reconstruction with limited angle in Dual Energy Breast CT

Limited angle Breast Computed Tomography uses lower energy and low projection angle to detect early breast cancer or other malignant tissues in the breast. The sensitivity of breast CT can be improved by applying dual energy technology. The general challenge which hampers full exploration of dual energy imaging is noise accumulation as a result of spectral overlaps from two different images. The author proposed hybrid optimization method (HOM) which leverages on fast convergence of simultaneous algebraic reconstruction techniques (SART) and good de-noising and artefacts removal of dictionary learning (DL) to minimizes noise in each image of dual energy and then apply decomposition on the noiseless dual data. The HOM algorithm is formulated as optimization problem which find good atoms from the dictionary obtained and dictionary atom are learned from training data set. The reconstructed images which are noise-free are then decomposed using DECT algorithm into two material basis. 2D phantom known as mbat-phantom consisting of two material basis (microcalcification and normal breast tissue) were simulated to test the algorithm. Noisy projection data were also simulated under the same condition by adding poison noise. The performance of the method was evaluated by estimating some image quality indices on reconstructed images and decomposed images. The proposed method shows the highest average structural similarity index map (SSIM) of 0.9987 and 0.9921 and peak signal to-noise ratio (PSNR) of 49.24 and 46.96 for reconstructed image without noise and noisy image respectively. Also, there is a reduction in average standard deviation (STD) error of decomposed image. Our method performs excellently in streak artefact removal and noise suppression which is capable of reconstructing faithful image in presence of noisy data.

Material decomposition for simulated dual-energy breast computed tomography via hybrid optimization method

BACKGROUND

Dual-energy breast CT reconstruction has a potential application that includes separation of microcalcification from healthy breast tissue for assisting early breast cancer detection.

OBJECTIVE

To investigate and validate the noise suppression algorithm applied in the decomposition of the simulated breast phantom into microcalcification and healthy breast.

METHODS

The proposed hybrid optimization method (HOM) uses a simultaneous algebraic reconstruction technique (SART) output as a prior image, which is then incorporated into the self-adaptive dictionary learning. This self-adaptive dictionary learning seeks each group of patches to faithfully represent the learned dictionary, and the sparsity and non-local similarity of group patches are used to enforce the image regularization term of the prior image. We simulate a numerical phantom by adding different levels of Gaussian noise to test performance of the proposed method.

RESULTS

The mean value of peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and root mean square error (RMSE) for the proposed method are (49.043±1.571), (0.997±0.002), (0.003±0.001) and (51.329±1.998), (0.998±0.002), (0.003±0.001) for 35 kVp and 49 kVp, respectively. The PSNR of the proposed method shows greater improvement over TWIST (5.2%), SART (34.6%), FBP (40.4%) and TWIST (3.7%), SART (39.9%), FBP (50.3%) for 35 kVp and 49 kVp energy images, respectively. For the proposed method, the signal-to-noise ratio (SNR) of decomposed normal breast tissue (NBT) is (22.036±1.535), which exceeded that of TWIST, SART, and FBP by 7.5%, 49.6%, and 96.4%, respectively. The results reveal that the proposed algorithm achieves the best performance in both reconstructed and decomposed images under different levels of noise and the performance is due to the high sparsity and good denoising ability of minimization exploited to solve the convex optimization problem.

CONCLUSIONS

This study demonstrates the potential of applying dual-energy reconstruction in breast CT to detect and separate clustered MCs from healthy breast tissues without noise amplification. Compared to other competing methods, the proposed algorithm achieves the best noise suppression performance for both reconstructed and decomposed images.

Study of role of digital breast tomosynthesis over digital mammography in the assessment of BIRADS 3 breast lesions

Background

Breast cancer is the most common malignancy in women and thus, screening has become an important health issue. Although mammography remains the standard of care for breast cancer screening and diagnosis (with biopsy), tomosynthesis (3D DBT) allows the separation of overlapping structures seen on 2D mammography and thus enables better depiction of masses or asymmetries.

Results

A prospective study for mammographic cases referred to our radiology unit included 60 lesions detected in 59 patients that were performed during the period from January 2016 to September 2017. Patients’ ages ranged from 26 to 72 years with mean age 51 ± 12 SD. Sixty percent of breast imaging-reporting and data system (BIRADS) 3 lesions detected by 2D digital mammography (36/60) changed their category after 3D DBT, 40% (24/60) digital mammography noticed lesions did not change their BIRADS after 3D DBT. Twenty-nine BIRADS 3 lesions out of the 60 were downgraded to BIRADS 1and 2, while 7 BIRADS 3 lesions out of the 60 were upgraded to BIRADS 4 and 5 which were all biopsied. Six out of the 7 lesions were pathologically proven ducal carcinoma and 1 out of 7 pathologically proven to be atypical ductal hyperplasia.

Conclusion

3D DBT significantly reduced the need for additional mammographic views and frequent follow-up studies as it gave better characterization for all BIRADS 3 lesions.

Initial Clinical Experience with Stationary Digital Breast Tomosynthesis

RATIONALE AND OBJECTIVES

A linear array of carbon nanotube-enabled x-ray sources allows for stationary digital breast tomosynthesis (sDBT), during which projection views are collected without the need to move the x-ray tube. This work presents our initial clinical experience with a first-generation sDBT device.

MATERIALS AND METHODS

Following informed consent, women with a "suspicious abnormality" (Breast Imaging Reporting and Data System 4), discovered by digital mammography and awaiting biopsy, were also imaged by the first generation sDBT. Four radiologists participated in this paired-image study, completing questionnaires while interpreting the mammograms and sDBT image stacks. Areas under the receiver operating characteristic curve were used to measure reader performance (likelihood of correctly identifying malignancy based on pathology as ground truth), while a multivariate analysis assessed preference, as readers compared one modality to the next when interpreting diagnostically important image features.

RESULTS

Findings from 43 women were available for analysis, in whom 12 cases of malignancy were identified by pathology. The mean areas under the receiver operating characteristic curve was significantly higher (p < 0.05) for sDBT than mammography for all breast density categories and breast thicknesses. Additionally, readers preferred sDBT over mammography when evaluating mass margins and shape, architectural distortion, and asymmetry, but preferred mammography when characterizing microcalcifications.

CONCLUSION

Readers preferred sDBT over mammography when interpreting soft-tissue breast features and were diagnostically more accurate using images generated by sDBT in a Breast Imaging Reporting and Data System 4 population. However, the findings also demonstrated the need to improve microcalcification conspicuity, which is guiding both technological and image-processing design changes in future sDBT devices.

Breast lesion size assessment in mastectomy specimens: Correlation of cone-beam breast-CT, digital breast tomosynthesis and full-field digital mammography with histopathology

To compare the accuracy of breast lesion size measurement of cone-beam breast-CT (CBBCT), digital breast tomosynthesis (DBT) and full-field digital mammography (FFDM).Patients scheduled for mastectomy due to at least 1 malignant breast lesion were included. Mastectomy specimens were examined by CBBCT, DBT, FFDM, and histopathology.A total of 94 lesions (40 patients) were included. Histopathological analyses revealed 47 malignant, 6 high-risk, and 41 benign lesions. Mean histopathological lesion size was 20.8 mm (range 2-100). Mean absolute size deviation from histopathology was largest for FFDM (5.3 ± 6.7 mm) and smallest for CBBCT 50 mA, high-resolution mode (4.3 ± 6.7 mm). Differences between imaging modalities did not reach statistical significance (P = .85).All imaging methods tend to overestimate breast lesion size compared to histopathological gold standard. No significant differences were found regarding size measurements, although in tendency CBBCT showed better lesion detection and cT classification over FFDM.

Lumpectomy Specimen Radiography: Does Orientation or 3-Dimensional Tomosynthesis Improve Margin Assessment?

PURPOSE

Our purpose was twofold. First, we sought to determine whether 2 orthogonal oriented views of excised breast cancer specimens could improve surgical margin assessment compared to a single unoriented view. Second, we sought to determine whether 3D tomosynthesis could improve surgical margin assessment compared to 2D mammography alone.

MATERIALS AND METHODS

Forty-one consecutive specimens were prospectively imaged using 4 protocols: single view unoriented 2D image acquired on a specimen unit (1VSU), 2 orthogonal oriented 2D images acquired on the specimen unit (2VSU), 2 orthogonal oriented 2D images acquired on a mammogram unit (2V2DMU), and 2 orthogonal oriented 3D images acquired on the mammogram unit (2V3DMU). Three breast imagers randomly assessed surgical margin of the 41 specimens with each protocol. Surgical margin per histopathology was considered the gold standard.

RESULTS

The average area under the curve (AUC) was 0.60 for 1VSU, 0.66 for 2VSU, 0.68 for 2V2DMU, and 0.60 for 2V3DMU. Comparing AUCs for 2VSU vs 1VSU by reader showed improved diagnostic accuracy using 2VSU; however, this difference was only statistically significant for reader 3 (0.73 vs 0.63, P = .0455). Comparing AUCs for 2V3DMU vs 2V2DMU by reader showed mixed results, with reader 1 demonstrating increased accuracy (0.72 vs 0.68, P = .5984), while readers 2 and 3 demonstrated decreased accuracy (0.50 vs 0.62, P = .1089 and 0.58 vs 0.75, P = .0269).

CONCLUSIONS

2VSU showed improved accuracy in surgical margin prediction compared to 1VSU, although this was not statistically significant for all readers. 3D tomosynthesis did not improve surgical margin assessment.

Visualizing microcalcifications in lumpectomy specimens: an exploration into the clinical potential of carbon nanotube-enabled stationary digital breast tomosynthesis

Purpose

To assess the visibility of microcalcifications in images generated by a first-generation carbon-nanotube (CNT)-enabled stationary digital breast tomosynthesis (sDBT) device, using magnified 2D mammography and conventional, moving-source DBT as references for comparison.

Methods

Lumpectomy specimens were imaged by magnified mammography and two 3D mammography approaches, including sDBT and moving-source DBT. The planar size of individual microcalcifications was measured in the reconstructed image stacks of sDBT and moving-source DBT and compared to the magnified mammography image. An artifact spread function (ASF) was used to assess the depth dimensions of the microcalcifications displayed through the reconstructed image stacks. Breast-imaging specialists rated their preference for one imaging modality over another when interpreting microcalcifications in the magnified mammography image and synthetic slab images from sDBT and moving-source DBT.

Results

The planar size of individual microcalcifications was similar in images generated by sDBT and moving-source DBT when the sDBT projections were binned to match the pixel size used by the moving-source DBT system. However, the unique structure of sDBT allowed for a wider-angle span of projection views and operation of the detector in full-resolution mode without significantly compromising the scan time. In this configuration, the planar sizes of individual microcalcifications displayed by sDBT was more similar to magnified mammography than moving-source DBT, and the microcalcifications had a narrower ASF through depth. Readers preferred sDBT over moving-source DBT when assessing microcalcifications in synthetic slab images, although magnified mammography was rated highest overall.

Conclusions

The sDBT system displayed microcalcifications as well as conventional, moving-source DBT when the effective pixel size of the detector was matched. However, with the detector in its full-resolution mode, sDBT displayed microcalcifications with greater clarity. Readers still preferred images generated by magnified mammography over both 3D mammography approaches. This finding is guiding continued hardware and software development to optimize the sDBT technology.

Why the Gold Standard Approach by Mammography Demands Extension by Multiomics? Application of Liquid Biopsy miRNA Profiles to Breast Cancer Disease Management

In the global context, the epidemic of breast cancer (BC) is evident for the early 21st century. Evidence shows that national mammography screening programs have sufficiently reduced BC related mortality. Therefore, the great utility of the mammography-based screening is not an issue. However, both false positive and false negative BC diagnosis, excessive biopsies, and irradiation linked to mammography application, as well as sub-optimal mammography-based screening, such as in the case of high-dense breast tissue in young females, altogether increase awareness among the experts regarding the limitations of mammography-based screening. Severe concerns regarding the mammography as the “golden standard” approach demanding complementary tools to cover the evident deficits led the authors to present innovative strategies, which would sufficiently improve the quality of the BC management and services to the patient. Contextually, this article provides insights into mammography deficits and current clinical data demonstrating the great potential of non-invasive diagnostic tools utilizing circulating miRNA profiles as an adjunct to conventional mammography for the population screening and personalization of BC management.

Using Google Trends Data to Study Public Interest in Breast Cancer Screening in Malaysia

Objective:

This study aims to investigate the public pattern in seeking breast cancer screening information in Malaysia using Google Trends.

Methods:

The Google Trends database was evaluated for the relative Internet search popularity of breast cancer and screening-related search terms from 2007 to 2018.

Results:

Result showed downward trends in breast cancer search, whereas mammogram and tomosynthesis search fluctuated consistently. A significant increment was found during Pink October month. Breast cancer search term achieved the highest popularity in the east coast of Malaysia with [x² (5, N=661) = 110.93, P<0.05], whereas mammogram attained the highest search volume in central Malaysia [x² (4, N=67) = 18.90, P<0.05]. The cross-correlation for breast cancer was moderate among northern Malaysia, Sabah, and Sarawak (0.3 ≤ r_s ≤ 0.7).

Conclusion:

Public interest trend in breast cancer screening is strongly correlated with the breast cancer awareness campaign, Pink October. Breast cancer screening should be promoted in the rural areas in Malaysia.

Computer-aided detection of mass in digital breast tomosynthesis using a faster region-based convolutional neural network

Digital breast tomosynthesis (DBT) is a newly developed three-dimensional tomographic imaging modality in the field of breast cancer screening designed to alleviate the limitations of conventional digital mammography-based breast screening methods. A computer-aided detection (CAD) system was designed for masses in DBT using a faster region-based convolutional neural network (faster-RCNN). To this end, a data set was collected, including 89 patients with 105 masses. An efficient detection architecture of convolution neural network with a region proposal network (RPN) was used for each slice to generate region proposals (i.e., bounding boxes) with a mass likelihood score. In each DBT volume, a slice fusion procedure was used to merge the detection results on consecutive 2D slices into one 3D DBT volume. The performance of the CAD system was evaluated using free-response receiver operating characteristic (FROC) curves. Our RCNN-based CAD system was compared with a deep convolutional neural network (DCNN)-based CAD system. The RCNN-based CAD generated a performance with an area under the ROC (AUC) of 0.96, whereas the DCNN-based CAD achieved a performance with AUC of 0.92. For lesion-based mass detection, the sensitivity of RCNN-based CAD was 90% at 1.54 false positive (FP) per volume, whereas the sensitivity of DCNN-based CAD was 90% at 2.81 FPs/volume. For breast-based mass detection, RCNN-based CAD generated a sensitivity of 90% at 0.76 FP/breast, which is significantly increased compared with the DCNN-based CAD with a sensitivity of 90% at 2.25 FPs/breast. The results suggest that the faster R-CNN has the potential to augment the prescreening and FP reduction in the CAD system for masses.

Calcifications at Digital Breast Tomosynthesis: Imaging Features and Biopsy Techniques

Full-field digital mammography (FFDM), the standard of care for breast cancer screening, has some limitations. With the advent of digital breast tomosynthesis (DBT), improvements including decreased recall rates and increased cancer detection rates have been observed. The quasi-three-dimensional capability of DBT reduces breast tissue overlap, a significant limitation of FFDM. However, early studies demonstrate that a few cancers detected at FFDM may not be diagnosed at DBT-only screening, and lesions with calcifications as the dominant feature may look less suspicious at DBT or not be visible at all. These findings support the use of combined FFDM and DBT protocols to optimize screening performance. However, this combination would approximately double the patient's radiation exposure. The development of computer algorithms that generate two-dimensional synthesized mammography (SM) views from DBT has improved calcification conspicuity and sensitivity. Therefore, SM may substitute for FFDM in screening protocols, reducing radiation exposure. DBT plus SM demonstrates significantly better performance than that of FFDM alone, although there are reports of missed malignant calcifications. Thus, some centers continue to perform FFDM with DBT. Use of DBT in breast imaging has also necessitated the development of DBT-guided biopsy. DBT-guided biopsy may have a higher success rate than that of stereotactic biopsy, with a shorter procedure time. While DBT brings substantial improvements to breast cancer imaging, it is important to be aware of its strengths and limitations regarding detection of calcifications. This article reviews the imaging appearance of breast calcifications at DBT, discusses calcification biopsy techniques, and provides an overview of the current literature. Online supplemental material is available for this article. ^©RSNA, 2019 An earlier incorrect version of this article appeared online. This article was corrected on February 13, 2019.

Detection of noncalcified breast cancer in patients with extremely dense breasts using digital breast tomosynthesis compared with full-field digital mammography

OBJECTIVE

To evaluate the tumour visibility and diagnostic performance of digital breast tomosynthesis (DBT) in patients with noncalcified T1 breast cancer.

METHODS

Medical records of 106 females with noncalcified T1 invasive breast cancer who underwent DBT and full-field digital mammography (FFDM) between January 2012 and December 2014 were retrospectively reviewed. To assess tumour visibility (score 1-3), all DBT and FFDM images were reviewed by two radiologists blinded to clinicopathological information. A reference standard was established by an unblinded consensus review of all images. Clinicopathological and imaging variables were analysed based on tumour visibility. After adding 159 negative controls, the diagnostic performance of DBT + FFDM was compared with that of FFDM.

RESULTS

The tumour visibility was significantly higher through DBT + FFDM (2.5 vs 1.8; p = 0.002) than FFDM alone. Breast composition was the independent variable for tumour visibility through DBT + FFDM (extremely dense; odds ratio, 0.02; p < 0.001). Sensitivity (p = 0.642), specificity (p = 0.463), positive-predictive value (p = 0.078), and negative-predictive value (p = 0.072) of DBT + FFDM were not significantly superior to those of FFDM in 55 females with extremely dense breast composition, whereas specificity (p = 0.002) and positive-predictive value (p < 0.001) were significantly higher in 210 females with other breast compositions.

CONCLUSION

Addition of DBT to FFDM showed no significant increase in the tumour visibility and diagnostic performance in patients with noncalcified T1 cancer in extremely dense breasts.

ADVANCES IN KNOWLEDGE

Addition of DBT to FFDM did not further improve the detection of noncalcified early breast cancers in females with extremely dense breasts.

Malignancy Detection on Mammography Using Dual Deep Convolutional Neural Networks and Genetically Discovered False Color Input Enhancement

Breast cancer is the most prevalent malignancy in the US and the third highest cause of cancer-related mortality worldwide. Regular mammography screening has been attributed with doubling the rate of early cancer detection over the past three decades, yet estimates of mammographic accuracy in the hands of experienced radiologists remain suboptimal with sensitivity ranging from 62 to 87% and specificity from 75 to 91%. Advances in machine learning (ML) in recent years have demonstrated capabilities of image analysis which often surpass those of human observers. Here we present two novel techniques to address inherent challenges in the application of ML to the domain of mammography. We describe the use of genetic search of image enhancement methods, leading us to the use of a novel form of false color enhancement through contrast limited adaptive histogram equalization (CLAHE), as a method to optimize mammographic feature representation. We also utilize dual deep convolutional neural networks at different scales, for classification of full mammogram images and derivative patches combined with a random forest gating network as a novel architectural solution capable of discerning malignancy with a specificity of 0.91 and a specificity of 0.80. To our knowledge, this represents the first automatic stand-alone mammography malignancy detection algorithm with sensitivity and specificity performance similar to that of expert radiologists.

Impact of Addition of Digital Breast Tomosynthesis to Digital Mammography in Lesion Characterization in Breast Cancer Patients

Context:

Digital breast tomosynthesis (DBT) is a new development in mammography technology which reduces the effect of overlapping tissue.

Aims:

The aim is to interrogate whether addition of DBT to digital mammography (DM) helps in better characterization of mammographic abnormalities in breast cancer patients in general and in different breast compositions.

Settings and Design:

Retrospective, analytical cross-sectional study.

Subjects and Methods:

Mammographic findings in 164 patients with 170 pathologically proven lesions were evaluated by using first DM alone and thereafter with addition of DBT to DM. The perceived utility of adjunct DBT was scored using a rating of 0–2. A score of 0 indicating that DM plus DBT was comparable to DM alone, 1 indicating that DM plus DBT was slightly better, and 2 indicating that DM plus DBT was definitely better.

Statistical Analysis:

McNemar Chi-squares test, Fisher's exact test.

Results:

On DM, 149 lesions were characterized mass with or without calcifications, 18 asymmetries with or without calcifications, 2 as architectural distortion, and 1 as microcalcification alone. Adjunct DBT helped in better morphological characterization of 17 lesions, with revelation of underlying masses in 16 asymmetries and one architectural distortion. Adjunct DBT was perceived to be slightly better than DM alone in 44.7% lesions, and definitely better in 22.9% lesions. Lesions showing score 1 or 2 improvement were significantly higher in heterogeneously and extremely dense breasts (P < 0.001).

Conclusions:

Adjunct DBT improves morphological characterization of lesions in patients with breast cancer. It highlights more suspicious features of lesions that indicate the presence of cancer, particularly in dense breasts.

An iterative reconstruction algorithm for digital breast tomosynthesis imaging using real data at three radiation doses

BACKGROUND

Iterative image reconstruction in Digital Breast Tomosynthesis (DBT) is a developing modality that produces three-dimensional (3D) reconstructed images of a breast to detect suspicious lesions. Algebraic reconstruction technique (ART), one of the iterative image reconstruction methods, was applied to reconstruct 3D data of breast and is becoming as one alternative method for the conventional image reconstruction techniques such as filtered back projection (FBP) in DBT imaging.

OBJECTIVE

A new majorization-minimization (MM) algorithm was presented for TV denoising of signals. In the field of DBT, however, the algorithm has not yet been applied. In this study, we proposed a new method of "ART+TV3D+MM," which applies (MM) algorithm to the images reconstructed by ART+TV3D for different imaging dose levels to investigate a possible reduction of radiation dose.

METHODS

Projections of a real breast phantom (CD Pasmam 1054) were acquired with a Siemens MAMMOMAT DBT system. The proposed new method was repeated and tested with 3 different radiation dose levels. The quality of the images reconstructed using the proposed new method were compared with those generated by the commonly used FBP method using both qualitative and quantitative assessments.

RESULTS

The new method showed superior results in terms of visual assessment, contrast to noise ratios (CNR), full width at half maximum (FWHM) values and 1D profiles compared with FBP of the Siemens MAMMOMAT. CNR values were evaluated for two different region of interests (ROIs). For instance, CNR values of ROI-2 of FBP and of new method were 1.670 and 1.978 at 100 mAs, respectively. Moreover, while CNR value of ROI-1 of FBP at 100 mAs was 0.955, CNR value of ROI-1 of using new method at 100 mAs was 48.163. FWHM values for FBP and the new method were 2.328 and 1.765 at 56 mAs, 2.032 and 1.661 at 100 mAs, and 2.111 and 1.736 at 199 mAs, respectively.

CONCLUSIONS

The results support that using the new method of "ART+TV3D+MM" could help decrease the radiation dose level, which is one of the most critical limitations of DBT imaging.

The Current Role of Image Compression Standards in Medical Imaging

With increasing utilization of medical imaging in clinical practice and the growing dimensions of data volumes generated by various medical imaging modalities, the distribution, storage, and management of digital medical image data sets requires data compression. Over the past few decades, several image compression standards have been proposed by international standardization organizations. This paper discusses the current status of these image compression standards in medical imaging applications together with some of the legal and regulatory issues surrounding the use of compression in medical settings.

Screening Mammography and Digital Breast Tomosynthesis: Controversies

Breast cancer screening with mammography reduces breast cancer mortality; however, diverging recommendations regarding screening have caused controversy. The emerging technology of digital breast tomosynthesis (DBT) may soon become the mainstay of breast cancer screening. We present recommendations for breast cancer screening based on guidelines. A PubMed literature review was performed and the results from five large clinical studies comparing the efficacy of digital mammography alone versus digital mammography with DBT are examined. We emphasize the importance of annual screening to reduce breast cancer mortality. Our review of the literature demonstrates that DBT increases cancer detection rates and reduces callbacks. Additional research is needed to determine whether the increased cancer detection rates are associated with a decrease in mortality.

Evidence on Synthesized Two-dimensional Mammography Versus Digital Mammography When Using Tomosynthesis (Three-dimensional Mammography) for Population Breast Cancer Screening

One limitation of using digital breast tomosynthesis (3-dimensional [3D] mammography) technology with conventional (2-dimensional [2D]) mammography for breast cancer (BC) screening is the increased radiation dose from dual acquisitions. To resolve this problem, synthesized 2D (s2D) reconstruction images similar to 2D mammography were developed using tomosynthesis acquisitions. The present review summarizes the evidence for s2D versus digital mammography (2D) when using tomosynthesis (3D) for BC screening to address whether using s2D instead of 2D (alongside 3D) will yield similar detection measures. Comparative population screening studies have provided consistent evidence that cancer detection rates do not differ between integrated 2D/3D (range, 5.45-8.5/1000 screens) and s2D/3D (range, 5.03-8.8/1000 screens). Also, although the recall measures were relatively heterogeneous across included studies, little difference was found between the 2 modalities. The mean glandular dose for s2D/3D was 55% to 58% of that for 2D/3D. In the context of BC screening, s2D/3D involves substantially less radiation than 2D/3D and provides similar detection measures. Thus, consideration of transitioning to tomosynthesis screening should aim to use s2D/3D to minimize harm.

Second generation stationary digital breast tomosynthesis system with faster scan time and wider angular span

PURPOSE

The aim of this study was to characterize a new generation stationary digital breast tomosynthesis system with higher tube flux and increased angular span over a first generation system.

METHODS

The linear CNT x-ray source was designed, built, and evaluated to determine its performance parameters. The second generation system was then constructed using the CNT x-ray source and a Hologic gantry. Upon construction, test objects and phantoms were used to characterize system resolution as measured by the modulation transfer function (MTF), and artifact spread function (ASF).

RESULTS

The results indicated that the linear CNT x-ray source was capable of stable operation at a tube potential of 49 kVp, and measured focal spot sizes showed source-to-source consistency with a nominal focal spot size of 1.1 mm. After construction, the second generation (Gen 2) system exhibited entrance surface air kerma rates two times greater the previous s-DBT system. System in-plane resolution as measured by the MTF is 7.7 cycles/mm, compared to 6.7 cycles/mm for the Gen 1 system. As expected, an increase in the z-axis depth resolution was observed, with a decrease in the ASF from 4.30 mm to 2.35 mm moving from the Gen 1 system to the Gen 2 system as result of an increased angular span.

CONCLUSIONS

The results indicate that the Gen 2 stationary digital breast tomosynthesis system, which has a larger angular span, increased entrance surface air kerma, and faster image acquisition time over the Gen 1 s-DBT system, results in higher resolution images. With the detector operating at full resolution, the Gen 2 s-DBT system can achieve an in-plane resolution of 7.7 cycles per mm, which is better than the current commercial DBT systems today, and may potentially result in better patient diagnosis.

Fully automated nipple detection in digital breast tomosynthesis

BACKGROUND AND OBJECTIVE

We propose a nipple detection algorithm for use with digital breast tomosynthesis (DBT) images. DBT images have been developed to overcome the weaknesses of 2D mammograms for denser breasts by providing 3D breast images. The nipple location acts as an invaluable landmark in DBT images for aligning the right and left breasts and describing the relative location of any existing lesions.

METHODS

Nipples may be visible or invisible in a breast image, and therefore a nipple detection method must be able to detect the nipples for both cases. The detection method for visible nipples based on their shape is simple and highly efficient. However, it is difficult to detect invisible nipples because they do not have a prominent shape. Fibroglandular tissue in a breast is anatomically connected with the nipple. Thus, the nipple location can be detected by analyzing the location of such tissue. In this paper, we propose a method for detecting the location of both visible and invisible nipples using fibroglandular tissue and changes in the breast area.

RESULTS

Our algorithm was applied to 138 DBT images, and its nipple detection accuracy was evaluated based on the mean Euclidean distance. The results indicate that our proposed method achieves a mean Euclidean distance of 3.10±2.58mm.

CONCLUSIONS

The nipple location can be a very important piece of information in the process of a DBT image registration. This paper presents a method for the automatic nipple detection in a DBT image. The extracted nipple location plays an essential role in classifying any existing lesions and comparing both the right and left breasts. Thus, the proposed method can help with computer-aided detection for a more efficient DBT image analysis.