Digital Database For Screening Mammography Research Articles

An effective fine grading method of BI-RADS classification in mammography.

Mammography is an important imaging technique for the detection of early breast cancer. Doctors classify mammograms as Breast Imaging Reporting and Data Systems (BI-RADS). This study aims to provide an intelligent BI-RADS grading prediction method, which can help radiologists and clinicians to distinguish the most challenging 4A, 4B, and 4C cases in mammography. Firstly, the breast region, the lesion region, and the corresponding region in the contralateral breast were extracted. Four categories of features were extracted from the original images and the images after the wavelet transform. Secondly, an optimized sequential forward floating selection (SFFS) was used for feature selection. Finally, a two-layer classifier integration was employed for fine grading prediction. 45 cases from the hospital and 500 cases from Digital Database for Screening Mammography (DDSM) database were used for evaluation. The classification performance of the support vector machine (SVM), Bayes, and random forest is very close on the 45 testing set, with the area under the receiver operating characteristic curve (AUC) of 0.978, 0.967, and 0.968. On the DDSM set, the AUC achieves 0.931, 0.938, and 0.874. Using the mean probability prediction, the AUC on the two datasets reaches 0.998 and 0.916. However, they are all significantly higher than the doctors' diagnosis, with the AUC of 0.807 and 0.725. A BI-RADS fine grading (2, 3, 4A, 4B, 4C, 5) prediction model was proposed. Through the evaluation from different datasets, the performance is proved higher than that of the doctors, which may provide great help for clinical BI-RADS classification diagnosis. Therefore, our method can produce more effective and reliable results.

Connected-UNets: a deep learning architecture for breast mass segmentation

Breast cancer analysis implies that radiologists inspect mammograms to detect suspicious breast lesions and identify mass tumors. Artificial intelligence techniques offer automatic systems for breast mass segmentation to assist radiologists in their diagnosis. With the rapid development of deep learning and its application to medical imaging challenges, UNet and its variations is one of the state-of-the-art models for medical image segmentation that showed promising performance on mammography. In this paper, we propose an architecture, called Connected-UNets, which connects two UNets using additional modified skip connections. We integrate Atrous Spatial Pyramid Pooling (ASPP) in the two standard UNets to emphasize the contextual information within the encoder–decoder network architecture. We also apply the proposed architecture on the Attention UNet (AUNet) and the Residual UNet (ResUNet). We evaluated the proposed architectures on two publically available datasets, the Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM) and INbreast, and additionally on a private dataset. Experiments were also conducted using additional synthetic data using the cycle-consistent Generative Adversarial Network (CycleGAN) model between two unpaired datasets to augment and enhance the images. Qualitative and quantitative results show that the proposed architecture can achieve better automatic mass segmentation with a high Dice score of 89.52%, 95.28%, and 95.88% and Intersection over Union (IoU) score of 80.02%, 91.03%, and 92.27%, respectively, on CBIS-DDSM, INbreast, and the private dataset.

Classification of mammographic breast density and its correlation with BI-RADS in elder women using machine learning approach

IntroductionMammographic breast density (MBD) is a known risk factor for breast cancer and older women have higher incidence rates of breast cancer occurrence. The Breast Imaging Reporting and Data System (BI-RADS) is a commonly used MBD classification tool for mammogram reporting. However, they have limitations since there are reading inconsistencies between different radiologists with the visual assessment of breast density. MethodsDigitised film-screen mammographic images were extracted from the Digital Database for Screening Mammography (DDSM). A machine learning project was developed using commercially available software with several predictive models applied to classify different amount of MBD on mammograms into different density groups. The effectiveness of different predictive models used in classifying the mammograms were tested by receiver operator characteristics (ROC) curve with comparison to the gold standard of BI-RADS classification. ResultsThree predictive models, Decision Tree (Tree), Support Vector Model (SVM) and k-Nearest Neighbour (kNN) showed high AUC values of 0.801, 0.805 and 0.810 respectively. High AUC values for the three predictive models indicates that the accuracy of the model is approaching that of the BI-RADS method. DiscussionOur machine learning project showed to have capabilities to be potentially used in the clinical settings to help categorise mammograms into extremely dense breasts (BI-RADS Group A) from entirely fatty breasts (BI-RADS Group D). ConclusionFindings from the present study suggest that the machine learning method is potentially useful to quantify the amount of MBD in mammograms.

Towards non-data-hungry and fully-automated diagnosis of breast cancer from mammographic images

Analysing local texture and generating features are two key issues for automatic cancer detection in mammographic images. Recent researches have shown that deep neural networks provide a promising alternative to hand-driven features which suffer from curse of dimensionality and low accuracy rates. However, large and balanced training data are foremost requirements for deep learning-based models and these data are not always available publicly. In this work, we propose a fully-automated method for breast cancer diagnosis that performs training using small sets of data. Feature extraction from mammographic images is performed using a genetic-programming-based descriptor that exploits statistics on a local binary pattern-like local distribution defined in each pixel. The effectiveness of the suggested method is demonstrated on two challenging datasets, (1) the digital database for screening mammography and (2) the mammographic image analysis society digital mammogram database, for content-based image retrieval as well as for abnormality/malignancy classification. The experimental results show that the proposed method outperforms or achieves comparable results with deep learning-based methods even those with transfer learning and/or data-augmentation.

DeepMiner: Discovering Interpretable Representations for Mammogram Classification and Explanation

We propose DeepMiner, a framework to discover interpretable representations in deep neural networks and to build explanations for medical predictions. By probing convolutional neural networks (CNNs) trained to classify cancer in mammograms, we show that many individual units in the final convolutional layer of a CNN respond strongly to diseased tissue concepts specified by the Breast Imaging-Reporting and Data System (BI-RADS) lexicon. After expert annotation of the interpretable units, our proposed method is able to generate explanations for CNN mammogram classification that are consistent with ground truth radiology reports on the Digital Database for Screening Mammography. We show that DeepMiner not only enables better understanding of the nuances of CNN classification decisions but also possibly discovers new visual knowledge relevant to medical diagnosis.

Segmentation of malignant tumours in mammogram images: A hybrid approach using convolutional neural networks and connected component analysis

AbstractThe segmentation of breast lesions is an important step in the computer‐aided analysis of the mammogram. The presence of noise in mammograms makes lesion detection challenging particularly for complex malignant lesions. Pre‐processing techniques can deal with the noise issue but distorts the important shape features. This motivates us to propose a novel hybrid approach by combining a convolution neural network (CNN) with connected component analysis (CCA) to segment malignant breast lesions without any pre‐processing to avoid any distortion in image sharpness at the initial stages. Two well‐known segmentation techniques namely, K‐means (KM) and Fuzzy c‐means (FCM) are also used to compare the results. From a pool of 1045 mammographic cancer images acquired from the Digital Database for Screening Mammography (DDSM), 1016 are used for training and validation, and 29 are used for testing. All three results (Hybrid, KM and FCM) are compared against the results by the expert Radiologist. The results indicate that, among various segmentation techniques, the proposed hybrid approach achieves the highest accuracy (90%), Matthew's correlation coefficient (0.79), Jaccard index (0.73) and the Dice similarity coefficient (0.84). Other performance evaluation techniques such as; precision, sensitivity, specificity, false‐positive rate, false discovery rate, negative predictive value and false‐negative rate also show the superior performance of the proposed hybrid approach. Statistical analysis (Mann–Whitney U test, T‐test, Chi‐square test, Kolmogorov–Smirnov test and Wilcoxon test), graphical analysis (Regression and Bland–Altman plots) and receiver operating characteristic curve further demonstrate the stability and consistency of the results.

Improvement of image enhancement for mammogram images using Fuzzy Anisotropic Diffusion Histogram Equalisation Contrast Adaptive Limited (FADHECAL)

ABSTRACT Image enhancement techniques for digital mammograms can preserve the contrast to distinguish the diagnostic features such as masses and microcalcifications. However, mammogram images suffer low contrast and high image noise due to low exposure radiation used. As a result, the diagnostic features are difficult to detect and analyse by the radiologist. In this article, we propose a novel Fuzzy Anisotropic Diffusion Histogram Equalisation Contrast Adaptive Limited (FADHECAL) enhancement technique to reduce the noise of the mammogram images while preserving contrast and brightness. The FADHECAL technique also applies Fuzzy Clipped Inference System (FCIS) which automatically select clip-limit during the enhancement process. The mammogram images were retrieved from Mini-Mammographic Image Analysis Society (MIAS) and Digital Database for Screening Mammography (DDSM) from the University of South Florida database. The results have shown that FADHECAL has the most superior results among other selected enhancement techniques with 6.502 ± 1.855 of AMBE, 0.934 ± 0.034 of SSIM, 15.742 ± 1.217 of MAE, 26.843 ± 2.541 of PSNR, 0.969 ± 0.021 of UIQI and 1.151 ± 0.147 of RMSE values. This FADHECAL can be used as an ideal platform of image enhancement for mammogram images to detect the breast cancer lesions with better noise reduction and preserving the image details.

BASCNet: Bilateral adaptive spatial and channel attention network for breast density classification in the mammogram

Breast density is a significant element for breast cancer precaution. The existing mammographic density classification methods cannot achieve satisfactory classification accuracy while achieving end-to-end. In this paper, we present a novel bilateral adaptive spatial and channel attention network (BASCNet) which integrates the information of the left and right breasts and adaptively pays attention to the discriminative features in spatial and channel dimensions. The proposed BASCNet has been fully proved on the public Digital Database for Screening Mammography (DDSM) and INbreast dataset, and the classification accuracies of 85.10% and 90.51% were achieved with fivefold cross-validation, respectively. Our method is fully automatic and has achieved the classification performance superior to the existing breast density classification methods. Massive ablation experiments were conducted to demonstrate the effectiveness of the network structure. Moreover, we compared the effects of different views (CC and MLO) on breast density classification and verified the effectiveness of the contralateral breast information integration. Overall, the proposed BASCNet has the potential to be applied to clinical diagnosis.

MommiNet-v2: Mammographic multi-view mass identification networks

Many existing approaches for mammogram analysis are based on single view. Some recent DNN-based multi-view approaches can perform either bilateral or ipsilateral analysis, while in practice, radiologists use both to achieve the best clinical outcome. MommiNet is the first DNN-based tri-view mass identification approach, which can simultaneously perform bilateral and ipsilateral analysis of mammographic images, and in turn, can fully emulate the radiologists’ reading practice. In this paper, we present MommiNet-v2, with improved network architecture and performance. Novel high-resolution network (HRNet)-based architectures are proposed to learn the symmetry and geometry constraints, to fully aggregate the information from all views for accurate mass detection. A multi-task learning scheme is adopted to incorporate both Breast Imaging-Reporting and Data System (BI-RADS) and biopsy information to train a mass malignancy classification network. Extensive experiments have been conducted on the public DDSM (Digital Database for Screening Mammography) dataset and our in-house dataset, and state-of-the-art results have been achieved in terms of mass detection accuracy. Satisfactory mass malignancy classification result has also been obtained on our in-house dataset.

Deep adversarial domain adaptation for breast cancer screening from mammograms

The early detection of breast cancer greatly increases the chances that the right decision for a successful treatment plan will be made. Deep learning approaches are used in breast cancer screening and have achieved promising results when a large-scale labeled dataset is available for training. However, they may suffer from a dramatic decrease in performance when annotated data are limited. In this paper, we propose a method called deep adversarial domain adaptation (DADA) to improve the performance of breast cancer screening using mammography. Specifically, our aim is to extract the knowledge from a public dataset (source domain) and transfer the learned knowledge to improve the detection performance on the target dataset (target domain). Because of the different distributions of the source and target domains, the proposed method adopts an adversarial learning technique to perform domain adaptation using the two domains. Specifically, the adversarial procedure is trained by taking advantage of the disagreement of two classifiers. To evaluate the proposed method, the public well-labeled image-level dataset Curated Breast Imaging Subset of the Digital Database for Screening Mammography (CBIS-DDSM) is employed as the source domain. Mammography samples from the West China Hospital were collected to construct our target domain dataset, and the samples are annotated at case-level based on the corresponding pathological reports. The experimental results demonstrate the effectiveness of the proposed method compared with several other state-of-the-art automatic breast cancer screening approaches.

Multi-scale attention-based convolutional neural network for classification of breast masses in mammograms.

Breast cancer is the cancer with the highest incidence in women, and early detection can effectively improve the survival rate of patients. Mammography is an important method for physicians to screening breast cancer, but the diagnosis of mammograms by physicians depends largely on clinical practice experience. Studies have shown that using computer-aided diagnosis techniques can help doctors diagnose breast cancer. In this paper, the method of convolutional neural network is mainly used to classify benign and malignant breast masses in the mammograms. First, we use multi-scale residual networks and densely connected networks as backbone networks to extract the features of global image patches and local image patches. Second, we use the attention module named convolutional block attention module (CBAM) to improve the two feature extraction networks to enhance the network's feature expression ability. Finally, we fuse the features of multi-scale image patches to achieve the classification of benign and malignant breast masses. In the digital database for screening mammography (DDSM) database, the accuracy, sensitivity, AUC value and corresponding standard deviation of our method are 0.9626±0.0110, 0.9719±0.0126, and 0.9576±0.0064, respectively. Compared with the commonly used ResNet (AUC=0.8823±0.0112) and DenseNet (AUC=0.9141±0.0085), the performance of our method has improved. In addition, we also used the INbreast database to train and validate the proposed method. The accuracy, sensitivity, AUC and corresponding standard deviations are 0.9554±0.0296, 0.9605±0.0228, and 0.9468±0.0085, respectively. Compared with the previous work, our proposed method uses multi-scale image features, has better classification performance in breast mass patches classification tasks, and can effectively assist physicians in breast cancer diagnosis.

Detection of Architectural Distorted Region in Mammogram Images Using Hybrid Classification Approaches

The architectural distorted regions in mammogram images are detected and segmented using computer aided hybrid classification approach in this paper. The main importance of this research work is to provide a computer aided methodology for screening the distorted regions in mammogram images. In present approach, the classification accuracy of the conventional methods is not suitable for further diagnosis process such as malignant and benign. Hence, the main objective of this paper is to develop an efficient architectural region detection method using soft computing method with high classification accuracy for further diagnosis purpose. This proposed method has two stages of the proposed flow as architectural distorted detected mammogram image and segmentation of architectural distorted regions in mammogram images. The first stage of this proposed method uses Random Forest (RF) classification method which classifies the source mammogram image into either normal or abnormal. In second stage of the proposed method, the abnormal image is further classified into either Benign or Malignant using Adaptive Neuro Fuzzy Inference System (ANFIS) classification approach. The proposed methodology for architectural distorted region detection is tested on the publicly available mammogram datasets Mammographic Image Analysis Society (MIAS) and Digital Database for Screening Mammography (DDSM) respectively. In this paper, the mammogram images from MIAS dataset are grouped into normal case (156 images), benign case (122 images) and malignant case (98 images). The mammogram images from DDSM dataset are grouped into normal case (144 images), benign case (112 images) and malignant case (145 images). The overall average detection rate of the proposed system on the mammogram images in MIAS dataset is about 98.7%. The overall average detection rate of the proposed system on the mammogram images in DDSM dataset is about 98.3%. The extensive simulations are carried out on the mammogram images which are obtained from these dataset and the results are compared with stated of art methods.

Detection of microcalcification clusters regions in mammograms combining discriminative deep belief networks

In order to overcome the shortcomings of high false positive rate and poor generalization in the detection of microcalcification clusters regions, this paper proposes a method combining discriminative deep belief networks (DDBNs) to automatically and quickly locate the regions of microcalcification clusters in mammograms. Firstly, the breast region was extracted and enhanced, and the enhanced breast region was segmented to overlapped sub-blocks. Then the sub-block was subjected to wavelet filtering. After that, DDBNs model for breast sub-block feature extraction and classification was constructed, and the pre-trained DDBNs was converted to deep neural networks (DNN) using a softmax classifier, and the network is fine-tuned by back propagation. Finally, the undetected mammogram was inputted to complete the location of suspicious lesions. By experimentally verifying 105 mammograms with microcalcifications from the Digital Database for Screening Mammography (DDSM), the method obtained a true positive rate of 99.45% and a false positive rate of 1.89%, and it only took about 16 s to detect a 2 888 × 4 680 image. The experimental results showed that the algorithm of this paper effectively reduced the false positive rate while ensuring a high positive rate. The detection of calcification clusters was highly consistent with expert marks, which provides a new research idea for the automatic detection of microcalcification clusters area in mammograms.

Breast Mass Detection in Mammography Based on Image Template Matching and CNN.

In recent years, computer vision technology has been widely used in the field of medical image processing. However, there is still a big gap between the existing breast mass detection methods and the real-world application due to the limited detection accuracy. It is known that humans locate the regions of interest quickly and further identify whether these regions are the targets we found. In breast cancer diagnosis, we locate all the potential regions of breast mass by glancing at the mammographic image from top to bottom and from left to right, then further identify whether these regions are a breast mass. Inspired by the process of human detection of breast mass, we proposed a novel breast mass detection method to detect breast mass on a mammographic image by stimulating the process of human detection. The proposed method preprocesses the mammographic image via the mathematical morphology method and locates the suspected regions of breast mass by the image template matching method. Then, it obtains the regions of breast mass by classifying these suspected regions into breast mass and background categories using a convolutional neural network (CNN). The bounding box of breast mass obtained by the mathematical morphology method and image template matching method are roughly due to the mathematical morphology method, which transforms all of the brighter regions into approximate circular areas. For regression of a breast mass bounding box, the optimal solution should be searched in the feasible region and the Particle Swarm Optimization (PSO) is suitable for solving the problem of searching the optimal solution within a certain range. Therefore, we refine the bounding box of breast mass by the PSO algorithm. The proposed breast mass detection method and the compared detection methods were evaluated on the open database Digital Database for Screening Mammography (DDSM). The experimental results demonstrate that the proposed method is superior to all of the compared detection methods in detection performance.

Automated early breast cancer detection and classification system

Early detection of breast cancer is clinically important to reduce the mortality rate. In this study, a new computer-aided detection (CAD) and classification system is introduced to classify two types of mammogram tumors (i.e., mass and calcification) as either benign or malignant. In this CAD system, the tumor-like regions (TLRs) are identified using the automated optimal Otsu thresholding method. Afterward, deep convolutional neural networks (CNNs) process the extracted TLRs to extract relevant mammogram features, investigating AlexNet and ResNet-50 architectures. The normalized extracted CNN features are further input to a support vector machine classifier to decode the classes of mammogram structures (i.e., Benign Calcification, Benign Mass, Malignant Calcification, and Malignant Mass nodules). The experimental results are tested on 2800 mammogram images from the Curated Breast Imaging Subset of Digital Database of Screening Mammography, a publicly available dataset. The accuracy of the proposed CAD system, to classify the ROI into one of the four classes, achieves 0.91 using AlexNet and 0.84 using ResNet-50 models, using fivefold cross-validation. Comparison results with the related methods confirm the advantages of the proposed CAD system.

Deep learning in mammography images segmentation and classification: Automated CNN approach

In this work, a new framework for breast cancer image segmentation and classification is proposed. Different models including InceptionV3, DenseNet121, ResNet50, VGG16 and MobileNetV2 models, are applied to classify Mammographic Image Analysis Society (MIAS), Digital Database for Screening Mammography (DDSM) and the Curated Breast Imaging Subset of DDSM (CBIS-DDSM) into benign and malignant. Moreover, the trained modified U-Net model is utilized to segment breast area from the mammogram images. This method will aid as a radiologist's assistant in early detection and improve the efficiency of our system. The Cranio Caudal (CC) vision and Mediolateral Oblique (MLO) view are widely used for the identification and diagnosis of breast cancer. The accuracy of breast cancer diagnosis will be improved as the number of views is increased. Our proposed frame work is based on MLO view and CC view to enhance the system performance. In addition, the lack of tagged data is a big challenge. Transfer learning and data augmentation are applied to overcome this problem. Three mammographic datasets; MIAS, DDSM and CBIS-DDSM, are utilized in our evaluation. End-to-end fully convolutional neural networks (CNNs) are introduced in this paper. The proposed technique of applying data augmentation with modified U-Net model and InceptionV3 achieves the best result, specifically with the DDSM dataset. This achieves 98.87% accuracy, 98.88% area under the curve (AUC), 98.98% sensitivity, 98.79% precision, 97.99% F1 score, and a computational time of 1.2134 s on DDSM datasets.

Computer-Assisted Diagnosis System for Abnormalities Classification in Digital Mammography Based on Multi-Threshold Modified Local Ternary Pattern (MtMLTP)

The aim of this paper is to develop an efficient breast cancer Computer Aided Diagnosis (CAD) system allowing the analysis of different breast tissues in mammograms and performing textural classification (normal, mass or microcalcification). Although several feature extraction algorithms for breast tissues analysis have been used, the findings concerning tissue characterization show no consensus in the literature. Specifically, the challenge may be great for mass and microcalcification detection on dense breasts. The proposed system is based on the development of a new feature extraction approach, the latter is called Multi-threshold Modified Local Ternary Pattern (MtMLTP), it allows the discrimination between various tissues in mammographic images allowing significant improvements in breast cancer diagnosis. In this paper, we have used 1000 ROIs obtained from Digital Database for Screening Mammography (DDSM) database and 100 ROIs from a local Tunisian database named Tunisian Digital Database for Screening Mammography (TDDSM). The Artificial Neural Network (ANN) shows good performance in the classification of abnormalities since the Area Under the Curve (AUC) of the proposed system has been found to be 0.97 for the DDSM database and 0.99 for the TDDSM Database.

Breast Cancer Mass Detection in Mammograms Using Gray Difference Weight and MSER Detector

Breast cancer is a deadly and one of the most prevalent cancers in women across the globe. Mammography is widely used imaging modality for diagnosis and screening of breast cancer. Segmentation of breast region and mass detection are crucial steps in automatic breast cancer detection. Due to the non-uniform distribution of various tissues, it is a challenging task to analyze mammographic images with high accuracy. In this paper, background suppression and pectoral muscle removal are performed using gradient weight map followed by gray difference weight and fast marching method. Enhancement of breast region is performed using contrast limited adaptive histogram equalization (CLAHE) and de-correlation stretch. Detection of breast masses is accomplished by gray difference weight and maximally stable external regions (MSER) detector. Experimentation on Mammographic Image Analysis Society (MIAS) and curated breast imaging subset of digital database for screening mammography (CBIS-DDSM) show that the method proposed performs breast boundary segmentation and mass detection with best accuracies. Mass detection achieved high accuracies of about 97.64% and 94.66% for MIAS and CBIS-DDSM dataset, respectively. The method is simple, robust, less affected to noise, density, shape and size which could provide reasonable support for mammographic analysis.

A framework for breast cancer classification using Multi-DCNNs

BackgroundDeep learning (DL) is the fastest-growing field of machine learning (ML). Deep convolutional neural networks (DCNN) are currently the main tool used for image analysis and classification purposes. There are several DCNN architectures among them AlexNet, GoogleNet, and residual networks (ResNet). MethodThis paper presents a new computer-aided diagnosis (CAD) system based on feature extraction and classification using DL techniques to help radiologists to classify breast cancer lesions in mammograms. This is performed by four different experiments to determine the optimum approach. The first one consists of end-to-end pre-trained fine-tuned DCNN networks. In the second one, the deep features of the DCNNs are extracted and fed to a support vector machine (SVM) classifier with different kernel functions. The third experiment performs deep features fusion to demonstrate that combining deep features will enhance the accuracy of the SVM classifiers. Finally, in the fourth experiment, principal component analysis (PCA) is introduced to reduce the large feature vector produced in feature fusion and to decrease the computational cost. The experiments are performed on two datasets (1) the curated breast imaging subset of the digital database for screening mammography (CBIS-DDSM) and (2) the mammographic image analysis society digital mammogram database (MIAS). ResultsThe accuracy achieved using deep features fusion for both datasets proved to be the highest compared to the state-of-the-art CAD systems. Conversely, when applying the PCA on the feature fusion sets, the accuracy did not improve; however, the computational cost decreased as the execution time decreased.

DeepCAT: Deep Computer-Aided Triage of Screening Mammography.

Although much deep learning research has focused on mammographic detection of breast cancer, relatively little attention has been paid to mammography triage for radiologist review. The purpose of this study was to develop and test DeepCAT, a deep learning system for mammography triage based on suspicion of cancer. Specifically, we evaluate DeepCAT's ability to provide two augmentations to radiologists: (1) discarding images unlikely to have cancer from radiologist review and (2) prioritization of images likely to contain cancer. We used 1878 2D-mammographic images (CC & MLO) from the Digital Database for Screening Mammography to develop DeepCAT, a deep learning triage system composed of 2 components: (1) mammogram classifier cascade and (2) mass detector, which are combined to generate an overall priority score. This priority score is used to order images for radiologist review. Of 595 testing images, DeepCAT recommended low priority for 315 images (53%), of which none contained a malignant mass. In evaluation of prioritizing images according to likelihood of containing cancer, DeepCAT's study ordering required an average of 26 adjacent swaps to obtain perfect review order. Our results suggest that DeepCAT could substantially increase efficiency for breast imagers and effectively triage review of mammograms with malignant masses.