Association of computerized mammographic parenchymal pattern measure with breast cancer risk: a pilot case-control study

Mammography classification with multi-view deep learning techniques: Investigating graph and transformer-based architectures

The potential and promise of deep learning systems to provide an independent assessment and relieve radiologists' burden in screening mammography have been recognized in several studies. However, the low cancer prevalence, the need to process high-resolution images, and the need to combine information from multiple views and scales still pose technical challenges. Multi-view architectures that combine information from the four mammographic views to produce an exam-level classification score are a promising approach to the automated processing of screening mammography. However, training such architectures from exam-level labels, without relying on pixel-level supervision, requires very large datasets and may result in suboptimal accuracy. Emerging architectures such as Visual Transformers (ViT) and graph-based architectures can potentially integrate ipsi-lateral and contra-lateral breast views better than traditional convolutional neural networks, thanks to their stronger ability of modeling long-range dependencies. In this paper, we extensively evaluate novel transformer-based and graph-based architectures against state-of-the-art multi-view convolutional neural networks, trained in a weakly-supervised setting on a middle-scale dataset, both in terms of performance and interpretability. Extensive experiments on the CSAW dataset suggest that, while transformer-based architecture outperform other architectures, different inductive biases lead to complementary strengths and weaknesses, as each architecture is sensitive to different signs and mammographic features. Hence, an ensemble of different architectures should be preferred over a winner-takes-all approach to achieve more accurate and robust results. Overall, the findings highlight the potential of a wide range of multi-view architectures for breast cancer classification, even in datasets of relatively modest size, although the detection of small lesions remains challenging without pixel-wise supervision or ad-hoc networks.

Mammographic Texture versus Conventional Cumulus Measure of Density in Breast Cancer Risk Prediction: A Literature Review

Mammographic textures show promise as breast cancer risk predictors, distinct from mammographic density. Yet, there is a lack of comprehensive evidence to determine the relative strengths as risk predictor of textures and density and the reliability of texture-based measures. We searched the PubMed database for research published up to November 2023, which assessed breast cancer risk associations [odds ratios (OR)] with texture-based measures and percent mammographic density (PMD), and their discrimination [area under the receiver operating characteristics curve (AUC)], using same datasets. Of 11 publications, for textures, six found stronger associations (P < 0.05) with 11% to 508% increases on the log scale by study, and four found weaker associations (P < 0.05) with 14% to 100% decreases, compared with PMD. Risk associations remained significant when fitting textures and PMD together. Eleven of 17 publications found greater AUCs for textures than PMD (P < 0.05); increases were 0.04 to 0.25 by study. Discrimination from PMD and these textures jointly was significantly higher than from PMD alone (P < 0.05). Therefore, different textures could capture distinct breast cancer risk information, partially independent of mammographic density, suggesting their joint role in breast cancer risk prediction. Some textures could outperform mammographic density for predicting breast cancer risk. However, obtaining reliable texture-based measures necessitates addressing various issues. Collaboration of researchers from diverse fields could be beneficial for advancing this complex field.

Deep Learning for Breast Cancer Risk Prediction: Application to a Large Representative UK Screening Cohort

Purpose To develop an artificial intelligence (AI) deep learning tool capable of predicting future breast cancer risk from a current negative screening mammographic examination and to evaluate the model on data from the UK National Health Service Breast Screening Program. Materials and Methods The OPTIMAM Mammography Imaging Database contains screening data, including mammograms and information on interval cancers, for more than 300 000 female patients who attended screening at three different sites in the United Kingdom from 2012 onward. Cancer-free screening examinations from women aged 50-70 years were performed and classified as risk-positive or risk-negative based on the occurrence of cancer within 3 years of the original examination. Examinations with confirmed cancer and images containing implants were excluded. From the resulting 5264 risk-positive and 191 488 risk-negative examinations, training (n = 89 285), validation (n = 2106), and test (n = 39 351) datasets were produced for model development and evaluation. The AI model was trained to predict future cancer occurrence based on screening mammograms and patient age. Performance was evaluated on the test dataset using the area under the receiver operating characteristic curve (AUC) and compared across subpopulations to assess potential biases. Interpretability of the model was explored, including with saliency maps. Results On the hold-out test set, the AI model achieved an overall AUC of 0.70 (95% CI: 0.69, 0.72). There was no evidence of a difference in performance across the three sites, between patient ethnicities, or across age groups. Visualization of saliency maps and sample images provided insights into the mammographic features associated with AI-predicted cancer risk. Conclusion The developed AI tool showed good performance on a multisite, United Kingdom-specific dataset. Keywords: Deep Learning, Artificial Intelligence, Breast Cancer, Screening, Risk Prediction Supplemental material is available for this article. ©RSNA, 2024.

Breast Density: Where Are We Now?

Breast density refers to the amount of fibroglandular tissue relative to fat on mammography and is determined either qualitatively through visual assessment or quantitatively. It is a heritable and dynamic trait associated with age, race/ethnicity, body mass index, and hormonal factors. Increased breast density has important clinical implications including the potential to mask malignancy and as an independent risk factor for the development of breast cancer. Breast density has been incorporated into breast cancer risk models. Given the impact of dense breasts on the interpretation of mammography, supplemental screening may be indicated.

Predicting Clinician Fixations on Glaucoma OCT Reports via CNN-Based Saliency Prediction Methods

Goal: To predict physician fixations specifically on ophthalmology optical coherence tomography (OCT) reports from eye tracking data using CNN based saliency prediction methods in order to aid in the education of ophthalmologists and ophthalmologists-in-training. Methods: Fifteen ophthalmologists were recruited to each examine 20 randomly selected OCT reports and evaluate the likelihood of glaucoma for each report on a scale of 0-100. Eye movements were collected using a Pupil Labs Core eye-tracker. Fixation heat maps were generated using fixation data. Results: A model trained with traditional saliency mapping resulted in a correlation coefficient (CC) value of 0.208, a Normalized Scanpath Saliency (NSS) value of 0.8172, a Kullback-Leibler (KLD) value of 2.573, and a Structural Similarity Index (SSIM) of 0.169. Conclusions: The TranSalNet model was able to predict fixations within certain regions of the OCT report with reasonable accuracy, but more data is needed to improve model accuracy. Future steps include increasing data collection, improving quality of data, and modifying the model architecture.

Predicting Breast Cancer Risk Using Radiomics Features of Mammography Images

Mammography images contain a lot of information about not only the mammary glands but also the skin, adipose tissue, and stroma, which may reflect the risk of developing breast cancer. We aimed to establish a method to predict breast cancer risk using radiomics features of mammography images and to enable further examinations and prophylactic treatment to reduce breast cancer mortality. We used mammography images of 4000 women with breast cancer and 1000 healthy women from the 'starting point set' of the OPTIMAM dataset, a public dataset. We trained a Light Gradient Boosting Machine using radiomics features extracted from mammography images of women with breast cancer (only the healthy side) and healthy women. This model was a binary classifier that could discriminate whether a given mammography image was of the contralateral side of women with breast cancer or not, and its performance was evaluated using five-fold cross-validation. The average area under the curve for five folds was 0.60122. Some radiomics features, such as 'wavelet-H_glcm_Correlation' and 'wavelet-H_firstorder_Maximum', showed distribution differences between the malignant and normal groups. Therefore, a single radiomics feature might reflect the breast cancer risk. The odds ratio of breast cancer incidence was 7.38 in women whose estimated malignancy probability was ≥0.95. Radiomics features from mammography images can help predict breast cancer risk.

Global Radiomic Features from Mammography for Predicting Difficult-To-Interpret Normal Cases

This work aimed to investigate whether global radiomic features (GRFs) from mammograms can predict difficult-to-interpret normal cases (NCs). Assessments from 537 readers interpreting 239 normal mammograms were used to categorise cases as 120 difficult-to-interpret and 119 easy-to-interpret based on cases having the highest and lowest difficulty scores, respectively. Using lattice- and squared-based approaches, 34 handcrafted GRFs per image were extracted and normalised. Three classifiers were constructed: (i) CC and (ii) MLO using the GRFs from corresponding craniocaudal and mediolateral oblique images only, based on the random forest technique for distinguishing difficult- from easy-to-interpret NCs, and (iii) CC + MLO using the median predictive scores from both CC and MLO models. Useful GRFs for the CC and MLO models were recognised using a scree test. The CC and MLO models were trained and validated using the leave-one-out-cross-validation. The models' performances were assessed by the AUC and compared using the DeLong test. A Kruskal-Wallis test was used to examine if the 34 GRFs differed between difficult- and easy-to-interpret NCs and if difficulty level based on the traditional breast density (BD) categories differed among 115 low-BD and 124 high-BD NCs. The CC + MLO model achieved higher performance (0.71 AUC) than the individual CC and MLO model alone (0.66 each), but statistically non-significant difference was found (all p > 0.05). Six GRFs were identified to be valuable in describing difficult-to-interpret NCs. Twenty features, when compared between difficult- and easy-to-interpret NCs, differed significantly (p < 0.05). No statistically significant difference was observed in difficulty between low- and high-BD NCs (p = 0.709). GRF mammographic analysis can predict difficult-to-interpret NCs.

Using Radiomics-Based Machine Learning to Create Targeted Test Sets to Improve Specific Mammography Reader Cohort Performance: A Feasibility Study

Mammography interpretation is challenging with high error rates. This study aims to reduce the errors in mammography reading by mapping diagnostic errors against global mammographic characteristics using a radiomics-based machine learning approach. A total of 36 radiologists from cohort A (n = 20) and cohort B (n = 16) read 60 high-density mammographic cases. Radiomic features were extracted from three regions of interest (ROIs), and random forest models were trained to predict diagnostic errors for each cohort. Performance was evaluated using sensitivity, specificity, accuracy, and AUC. The impact of ROI placement and normalization on prediction was investigated. Our approach successfully predicted both the false positive and false negative errors of both cohorts but did not consistently predict location errors. The errors produced by radiologists from cohort B were less predictable compared to those in cohort A. The performance of the models did not show significant improvement after feature normalization, despite the mammograms being produced by different vendors. Our novel radiomics-based machine learning pipeline focusing on global radiomic features could predict false positive and false negative errors. The proposed method can be used to develop group-tailored mammographic educational strategies to help improve future mammography reader performance.

Beyond Breast Density: Risk Measures for Breast Cancer in Multiple Imaging Modalities

Breast density is an independent risk factor for breast cancer. In digital mammography and digital breast tomosynthesis, breast density is assessed visually using the four-category scale developed by the American College of Radiology Breast Imaging Reporting and Data System (5th edition as of November 2022). Epidemiologically based risk models, such as the Tyrer-Cuzick model (version 8), demonstrate superior modeling performance when mammographic density is incorporated. Beyond just density, a separate mammographic measure of breast cancer risk is parenchymal textural complexity. With advancements in radiomics and deep learning, mammographic textural patterns can be assessed quantitatively and incorporated into risk models. Other supplemental screening modalities, such as breast US and MRI, offer independent risk measures complementary to those derived from mammography. Breast US allows the two components of fibroglandular tissue (stromal and glandular) to be visualized separately in a manner that is not possible with mammography. A higher glandular component at screening breast US is associated with higher risk. With MRI, a higher background parenchymal enhancement of the fibroglandular tissue has also emerged as an imaging marker for risk assessment. Imaging markers observed at mammography, US, and MRI are powerful tools in refining breast cancer risk prediction, beyond mammographic density alone.

Mammographic Breast Density: Current Assessment Methods, Clinical Implications, and Future Directions

Mammographic breast density is widely accepted as an independent risk factor for the development of breast cancer. In addition, because dense breast tissue may mask breast malignancies, breast density is inversely related to the sensitivity of screening mammography. Given the risks associated with breast density, as well as ongoing efforts to stratify individual risk and personalize breast cancer screening and prevention, numerous studies have sought to better understand the factors that impact breast density, and to develop and implement reproducible, quantitative methods to assess mammographic density. Breast density assessments have been incorporated into risk assessment models to improve risk stratification. Recently, novel techniques for analyzing mammographic parenchymal complexity, or texture, have been explored as potential means of refining mammographic tissue-based risk assessment beyond breast density.

Applying artificial intelligence technology to assist with breast cancer diagnosis and prognosis prediction

Breast cancer remains the most diagnosed cancer in women. Advances in medical imaging modalities and technologies have greatly aided in the early detection of breast cancer and the decline of patient mortality rates. However, reading and interpreting breast images remains difficult due to the high heterogeneity of breast tumors and fibro-glandular tissue, which results in lower cancer detection sensitivity and specificity and large inter-reader variability. In order to help overcome these clinical challenges, researchers have made great efforts to develop computer-aided detection and/or diagnosis (CAD) schemes of breast images to provide radiologists with decision-making support tools. Recent rapid advances in high throughput data analysis methods and artificial intelligence (AI) technologies, particularly radiomics and deep learning techniques, have led to an exponential increase in the development of new AI-based models of breast images that cover a broad range of application topics. In this review paper, we focus on reviewing recent advances in better understanding the association between radiomics features and tumor microenvironment and the progress in developing new AI-based quantitative image feature analysis models in three realms of breast cancer: predicting breast cancer risk, the likelihood of tumor malignancy, and tumor response to treatment. The outlook and three major challenges of applying new AI-based models of breast images to clinical practice are also discussed. Through this review we conclude that although developing new AI-based models of breast images has achieved significant progress and promising results, several obstacles to applying these new AI-based models to clinical practice remain. Therefore, more research effort is needed in future studies.

Radiomics in photon-counting dedicated breast CT: potential of texture analysis for breast density classification

Background

We investigated whether features derived from texture analysis (TA) can distinguish breast density (BD) in spiral photon-counting breast computed tomography (PC-BCT).

Methods

In this retrospective single-centre study, we analysed 10,000 images from 400 PC-BCT examinations of 200 patients. Images were categorised into four-level density scale (a–d) using Breast Imaging Reporting and Data System (BI-RADS)-like criteria. After manual definition of representative regions of interest, 19 texture features (TFs) were calculated to analyse the voxel grey-level distribution in the included image area. ANOVA, cluster analysis, and multinomial logistic regression statistics were used. A human readout then was performed on a subset of 60 images to evaluate the reliability of the proposed feature set.

Results

Of the 19 TFs, 4 first-order features and 7 second-order features showed significant correlation with BD and were selected for further analysis. Multinomial logistic regression revealed an overall accuracy of 80% for BD assessment. The majority of TFs systematically increased or decreased with BD. Skewness (rho -0.81), as a first-order feature, and grey-level nonuniformity (GLN, -0.59), as a second-order feature, showed the strongest correlation with BD, independently of other TFs. Mean skewness and GLN decreased linearly from density a to d. Run-length nonuniformity (RLN), as a second-order feature, showed moderate correlation with BD, but resulted in redundant being correlated with GLN. All other TFs showed only weak correlation with BD (range -0.49 to 0.49, p < 0.001) and were neglected.

Conclusion

TA of PC-BCT images might be a useful approach to assess BD and may serve as an observer-independent tool.

Reducing False-Positive Biopsies using Deep Neural Networks that Utilize both Local and Global Image Context of Screening Mammograms

Breast cancer is the most common cancer in women, and hundreds of thousands of unnecessary biopsies are done around the world at a tremendous cost. It is crucial to reduce the rate of biopsies that turn out to be benign tissue. In this study, we build deep neural networks (DNNs) to classify biopsied lesions as being either malignant or benign, with the goal of using these networks as second readers serving radiologists to further reduce the number of false-positive findings. We enhance the performance of DNNs that are trained to learn from small image patches by integrating global context provided in the form of saliency maps learned from the entire image into their reasoning, similar to how radiologists consider global context when evaluating areas of interest. Our experiments are conducted on a dataset of 229,426 screening mammography examinations from 141,473 patients. We achieve an AUC of 0.8 on a test set consisting of 464 benign and 136 malignant lesions.

Algorithms and methods for computerized analysis of mammography images in breast cancer risk assessment

BACKGROUND AND OBJECTIVES

The computerized analysis of mammograms for the development of quantitative biomarkers is a growing field with applications in breast cancer risk assessment. Computerized image analysis offers the possibility of using different methods and algorithms to extract additional information from screening and diagnosis images to aid in the assessment of breast cancer risk. In this work, we review the algorithms and methods for the automated, computerized analysis of mammography images for the task mentioned, and discuss the main challenges that the development and improvement of these methods face today.

METHODS

We review the recent progress in two main branches of mammography-based risk assessment: parenchymal analysis and breast density estimation, including performance indicators of most of the studies considered. Parenchymal analysis methods are divided into feature-based methods and deep learning-based methods; breast density methods are grouped into area-based, volume-based, and breast categorization methods. Additionally, we identify the challenges that these study fields currently face.

RESULTS

Parenchymal analysis using deep learning algorithms are on the rise, with some studies showing high-performance indicators, such as an area under the receiver operating characteristic curve of up to 90. Methods for risk assessment using breast density report a wider variety of performance indicators; however, we can also identify that the approaches using deep learning methods yield high performance in each of the subdivisions considered.

CONCLUSIONS

Both breast density estimation and parenchymal analysis are promising tools for the task of breast cancer risk assessment; deep learning methods have shown performance comparable or superior to the other considered methods. All methods considered face challenges such as the lack of objective comparison between them and the lack of access to datasets from different populations.

An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization

Medical images differ from natural images in significantly higher resolutions and smaller regions of interest. Because of these differences, neural network architectures that work well for natural images might not be applicable to medical image analysis. In this work, we propose a novel neural network model to address these unique properties of medical images. This model first uses a low-capacity, yet memory-efficient, network on the whole image to identify the most informative regions. It then applies another higher-capacity network to collect details from chosen regions. Finally, it employs a fusion module that aggregates global and local information to make a prediction. While existing methods often require lesion segmentation during training, our model is trained with only image-level labels and can generate pixel-level saliency maps indicating possible malignant findings. We apply the model to screening mammography interpretation: predicting the presence or absence of benign and malignant lesions. On the NYU Breast Cancer Screening Dataset, our model outperforms (AUC = 0.93) ResNet-34 and Faster R-CNN in classifying breasts with malignant findings. On the CBIS-DDSM dataset, our model achieves performance (AUC = 0.858) on par with state-of-the-art approaches. Compared to ResNet-34, our model is 4.1x faster for inference while using 78.4% less GPU memory. Furthermore, we demonstrate, in a reader study, that our model surpasses radiologist-level AUC by a margin of 0.11.

Association of the Differences in Average Glandular Dose with Breast Cancer Risk

Objectives. To compare the differences in normalized average glandular dose (NAGD) between the breasts of healthy subjects and those of cancer patients and to determine if the NAGD difference is associated with breast cancer risk and improves breast cancer classification. Materials and Methods. Craniocaudal view and mediolateral view full-field digital mammography (FFDM) images were obtained from 1682 healthy subjects whose breasts were categorized as Breast Imaging-Reporting and Data System (BI-RADS) I or II and from 811 biopsy-confirmed unilateral breast cancer patients whose breasts on the contralateral side were category I or II. Both populations were randomized into training and test sets. Multivariate logistic regression analysis was used to build the breast cancer risk assessment model, and the area under the receiver operating characteristic curve ( $A_{\text{z}}$ ) was used to evaluate the model. Twenty-two breast cancer patients who were originally categorized as BI-RADS I or II for both breasts, but were diagnosed with unilateral biopsy-confirmed breast cancer subsequently, were included to validate the model. Results. The NAGD differences in both FFDM images between tumor-bearing breasts and the healthy breasts of patients were significantly higher than those in healthy subjects ( $P<0.001$ ). The model with NAGD differences had a higher $A_{\text{z}}$ value than the model without NAGD differences. While there was no NAGD differences between originally healthy breasts of breast cancer patients, significant NAGD differences between now tumor-bearing breasts and the then previously healthy breasts were found in both FFDM images. Conclusions. NAGD differences between both breasts can be included in the breast cancer risk assessment model to evaluate breast cancer risk.

Additive value of exposure parameters for breast cancer diagnosis in digital mammography

OBJECTIVES

To investigate the effect of different breast lesions on exposure parameters in digital mammography and to determine whether the exposure parameters can additively improve diagnostic efficiency.

METHODS

Craniocaudal view and mediolateral view full-field digital mammography images from 982 women with unilateral lesions (341 with malignant lesions, 189 with benign lesions, and 452 healthy women) obtained at Nanfang Hospital were reviewed. Differences in exposure parameters (tube voltage and load, breast thickness (BT), and average glandular dose (AGD)) between breasts were calculated. The relationships between parameter differences and lesion size were explored. A logistic regression model was used based on the AGD and BT differences, and the area under the receiver operating characteristic curve (AUC) was used to assess the performance of these parameters in differentiating malignant from benign and healthy subjects. Independently, data from 129 women (82 with malignant and 47 with benign lesions) treated at Sun Yat-sen Memorial Hospital were collected to validate the model.

RESULTS

Differences in tube voltage and load, BT, and AGD between breasts were significantly greater in the malignant subjects than benign (p < 0.05) and healthy subjects (p < 0.05). The AUCs for the comparisons of malignant vs. healthy subjects, malignant vs. benign subjects, and benign vs. healthy subjects were 0.77 ± 0.02, 0.72 ± 0.02, and 0.57 ± 0.02, respectively. The model combining the exposure parameters with the BI-RADS category resulted in a higher AUC (0.910 ± 0.03) compared with physician diagnosis alone (0.820 ± 0.04) for differentiating between malignant and benign lesions.

CONCLUSIONS

Exposure parameters additively improved diagnostic accuracy for breast cancer and yielded more reliable results.

KEY POINTS

• Differences in kVp, mAs, BT, and AGD between breasts were significantly greater in the malignant subjects than benign and healthy subjects. • The model combining exposure parameters with the BI-RADS category resulted in a higher AUC compared with the physician's diagnosis for differentiating between malignant and benign lesions. • Exposure parameters additively improved diagnostic accuracy for breast cancer.

A Comparison of Regions of Interest in Parenchymal Analysis for Breast Cancer Risk Assessment

Computerized parenchymal analysis has shown potential to be utilized as an imaging biomarker to estimate the risk of breast cancer. Parenchymal analysis of digital mammograms is based on the extraction of computerized measures to build machine learning-based models for the prediction of breast cancer risk. However, the choice of the region of interest (ROI) for feature extraction within the breast remains an open problem. In this work we perform a comparison between five different methods suggested in the literature for automated ROI selection, including the whole breast (WB), the maximum squared (MS), the retro-areolar region (RA), the lattice-based (LB), and the polar-based (PB) selection methods. For the experiments, we built a retrospective dataset of 896 screening mammograms from 224 women (112 cases and 112 healthy controls). The performance of each ROI selection method was measured in terms of the area under the curve (AUC) values. The AUC values varied between 0.55 and 0.79 depending on the method and experimental settings. The best performance on an independent test set was achieved by the MS method (AUC of 0.59, 95% CI: 0.55-0.64). This method is fully-automated and does not require adjusting hyper-parameters. Based on our results, we prompt the use of the MS method for ROI selection in the computerized parenchymal analysis for breast cancer risk assessment.

Do Mammographic Systems Affect the Performance of Computerized Parenchymal Analysis?

Early identification of women at high risk of developing breast cancer is fundamental for timely diagnosis and treatment. Recently, researchers have demonstrated that the computerized analysis of parenchymal (breast tissue) patterns in mammograms can be utilized to assess the risk level of patients. However, parenchymal analysis being an image-based biomarker, its performance may be affected by the acquisition parameters of the mammogram. Unfortunately, research on the effect of the mammographic system on the performance of parenchymal analysis is very scarce. In this paper, we implement a parenchymal analysis algorithm and study the effect of different mammographic systems on its performance. We show in a setting of 286 women that the use of different mammographic systems can yield differences of up to 24% in the area under the ROC curve. Results suggest the the construction of models for risk assessment based on parenchymal analysis should incorporate the imaging technologies into the analysis.

Clinical evaluation of a fully-automated parenchymal analysis software for breast cancer risk assessment: A pilot study in a Finnish sample

PURPOSE

To assess the association between breast cancer risk and mammographic parenchymal measures obtained using a fully-automated, publicly available software, OpenBreast.

METHODS

This retrospective case-control study involved screening mammograms of asymptomatic women diagnosed with breast cancer between 2016 and 2017. The 114 cases were matched with corresponding healthy controls by birth and screening years and the mammographic system used. Parenchymal analysis was performed using OpenBreast, a software implementing a computerized parenchymal analysis algorithm. Breast percent density was measured with an interactive thresholding method. The parenchymal measures were Box-Cox transformed and adjusted for age and percent density. Changes in the odds ratio per standard deviation (OPERA) with 95% confidence intervals (CIs) and the area under the ROC curve (AUC) for parenchymal measures and percent densities were used to evaluate the discrimination between cases and controls. Differences in AUCs were assessed using DeLong's test.

RESULTS

The adjusted OPERA value of parenchymal measures was 2.49 (95% CI: 1.79-3.47). Parenchymal measures using OpenBreast were more accurate (AUC = 0.779) than percent density (AUC = 0.609) in discriminating between cases and controls (p < 0.001).

CONCLUSIONS

Parenchymal measures obtained with the evaluated software were positively associated with breast cancer risk and were more accurate than percent density in the prediction of risk.

Artificial intelligence in the interpretation of breast cancer on MRI

Advances in both imaging and computers have led to the rise in the potential use of artificial intelligence (AI) in various tasks in breast imaging, going beyond the current use in computer-aided detection to include diagnosis, prognosis, response to therapy, and risk assessment. The automated capabilities of AI offer the potential to enhance the diagnostic expertise of clinicians, including accurate demarcation of tumor volume, extraction of characteristic cancer phenotypes, translation of tumoral phenotype features to clinical genotype implications, and risk prediction. The combination of image-specific findings with the underlying genomic, pathologic, and clinical features is becoming of increasing value in breast cancer. The concurrent emergence of newer imaging techniques has provided radiologists with greater diagnostic tools and image datasets to analyze and interpret. Integrating an AI-based workflow within breast imaging enables the integration of multiple data streams into powerful multidisciplinary applications that may lead the path to personalized patient-specific medicine. In this article we describe the goals of AI in breast cancer imaging, in particular MRI, and review the literature as it relates to the current application, potential, and limitations in breast cancer. Level of Evidence: 3 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2020;51:1310-1324.

Convolutional Neural Network Based Breast Cancer Risk Stratification Using a Mammographic Dataset

RATIONALE AND OBJECTIVES

We propose a novel convolutional neural network derived pixel-wise breast cancer risk model using mammographic dataset.

MATERIALS AND METHODS

An institutional review board approved retrospective case-control study of 1474 mammographic images was performed in average risk women. First, 210 patients with new incidence of breast cancer were identified. Mammograms from these patients prior to developing breast cancer were identified and made up the case group [420 bilateral craniocaudal mammograms]. The control group consisted of 527 patients without breast cancer from the same time period. Prior mammograms from these patients made up the control group [1054 bilateral craniocaudal mammograms]. A convolutional neural network (CNN) architecture was designed for pixel-wise breast cancer risk prediction. Briefly, each mammogram was normalized as a map of z-scores and resized to an input image size of 256 × 256. Then a contracting and expanding fully convolutional CNN architecture was composed entirely of 3 × 3 convolutions, a total of four strided convolutions instead of pooling layers, and symmetric residual connections. L2 regularization and augmentation methods were implemented to prevent overfitting. Cases were separated into training (80%) and test sets (20%). A 5-fold cross validation was performed. Software code was written in Python using the TensorFlow module on a Linux workstation with NVIDIA GTX 1070 Pascal GPU.

RESULTS

The average age of patients between the case and the control groups was not statistically different [case: 57.4years (SD, 10.4) and control: 58.2years (SD, 10.9), p = 0.33]. Breast Density (BD) was significantly higher in the case group [2.39 (SD, 0.7)] than the control group [1.98 (SD, 0.75), p < 0.0001]. On multivariate logistic regression analysis, both CNN pixel-wise mammographic risk model and BD were significant independent predictors of breast cancer risk (p < 0.0001). The CNN risk model showed greater predictive potential [OR = 4.42 (95% CI, 3.4-5.7] compared to BD [OR = 1.67 (95% CI, 1.4-1.9). The CNN risk model achieved an overall accuracy of 72% (95%CI, 69.8-74.4) in predicting patients in the case group.

CONCLUSION

Novel pixel-wise mammographic breast evaluation using a CNN architecture can stratify breast cancer risk, independent of the BD. Larger dataset will likely improve our model.

Digital Mammography in Breast Cancer: Additive Value of Radiomics of Breast Parenchyma

Background Previous studies have suggested that breast parenchymal texture features may reflect the biologic risk factors associated with breast cancer development. Therefore, combining the characteristics of normal parenchyma from the contralateral breast with radiomic features of breast tumors may improve the accuracy of digital mammography in the diagnosis of breast cancer. Purpose To determine whether the addition of radiomic analysis of contralateral breast parenchyma to the characterization of breast lesions with digital mammography improves lesion classification over that with radiomic tumor features alone. Materials and Methods This HIPAA-compliant, retrospective study included 182 patients (age range, 25-90 years; mean age, 55.9 years ± 14.9) who underwent mammography between June 2002 and July 2009. There were 106 malignant and 76 benign lesions. Automatic lesion segmentation and radiomic analysis were performed for each breast lesion. Radiomic texture analysis was applied in the normal regions of interest in the contralateral breast parenchyma to assess the mammographic parenchymal patterns. The classification performance of both individual features and the output from a Bayesian artificial neural network classifier was evaluated with the leave-one-patient-out method by using the area under the receiver operating characteristic curve (AUC) as the figure of merit in the task of differentiating between malignant and benign lesions. Results The performance of the combined lesion and parenchyma classifier in the differentiation between malignant and benign mammographic lesions was better than that with the lesion features alone (AUC = 0.84 ± 0.03 vs 0.79 ± 0.03, respectively; P = .047). Overall, six radiomic features-spiculation, margin sharpness, size, circularity from the tumor feature set, and skewness and power law beta from the parenchymal feature set-were selected more than 50% of the time during the feature selection process on the combined feature set. Conclusion Combining quantitative radiomic data from tumors with contralateral parenchyma characterizations may improve diagnostic accuracy for breast cancer. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Shaffer in this issue.

Incorporating Breast Anatomy in Computational Phenotyping of Mammographic Parenchymal Patterns for Breast Cancer Risk Estimation

We retrospectively analyzed negative screening digital mammograms from 115 women who developed unilateral breast cancer at least one year later and 460 matched controls. Texture features were estimated in multiple breast regions defined by an anatomically-oriented polar grid, and were weighted by their position and underlying dense versus fatty tissue composition. Elastic net regression with cross-validation was performed and area under the curve (AUC) of the receiver operating characteristic (ROC) was used to evaluate ability to predict breast cancer. We also compared our anatomy-augmented features to current state-of-the-art in which parenchymal texture was assessed without considering breast anatomy and evaluated the added value of the extracted features to breast density, body-mass-index (BMI) and age as baseline predictors. Our anatomy-augmented texture features resulted in higher discriminatory capacity (AUC = 0.63 vs. AUC = 0.59) when breast anatomy was not considered (p = 0.021), with dense tissue regions and the central breast quadrant being more heavily weighted. Texture also improved baseline models (from AUC = 0.62 to AUC = 0.67, p = 0.029). Our findings suggest that incorporating breast anatomy information could augment imaging markers of breast cancer risk with the potential to improve personalized breast cancer risk assessment.

Radiomic Phenotypes of Mammographic Parenchymal Complexity: Toward Augmenting Breast Density in Breast Cancer Risk Assessment

Purpose To identify phenotypes of mammographic parenchymal complexity by using radiomic features and to evaluate their associations with breast density and other breast cancer risk factors. Materials and Methods Computerized image analysis was used to quantify breast density and extract parenchymal texture features in a cross-sectional sample of women screened with digital mammography from September 1, 2012, to February 28, 2013 (n = 2029; age range, 35-75 years; mean age, 55.9 years). Unsupervised clustering was applied to identify and reproduce phenotypes of parenchymal complexity in separate training (n = 1339) and test sets (n = 690). Differences across phenotypes by age, body mass index, breast density, and estimated breast cancer risk were assessed by using Fisher exact, χ², and Kruskal-Wallis tests. Conditional logistic regression was used to evaluate preliminary associations between the detected phenotypes and breast cancer in an independent case-control sample (76 women diagnosed with breast cancer and 158 control participants) matched on age. Results Unsupervised clustering in the screening sample identified four phenotypes with increasing parenchymal complexity that were reproducible between training and test sets (P = .001). Breast density was not strongly correlated with phenotype category (R² = 0.24 for linear trend). The low- to intermediate-complexity phenotype (prevalence, 390 of 2029 [19%]) had the lowest proportion of dense breasts (eight of 390 [2.1%]), whereas similar proportions were observed across other phenotypes (from 140 of 291 [48.1%] in the high-complexity phenotype to 275 of 511 [53.8%] in the low-complexity phenotype). In the independent case-control sample, phenotypes showed a significant association with breast cancer (P = .001), resulting in higher discriminatory capacity when added to a model with breast density and body mass index (area under the curve, 0.84 vs 0.80; P = .03 for comparison). Conclusion Radiomic phenotypes capture mammographic parenchymal complexity beyond conventional breast density measures and established breast cancer risk factors. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Pinker in this issue.

Applying a new computer-aided detection scheme generated imaging marker to predict short-term breast cancer risk

This study aims to investigate the feasibility of identifying a new quantitative imaging marker based on false-positives generated by a computer-aided detection (CAD) scheme to help predict short-term breast cancer risk. An image dataset including four view mammograms acquired from 1044 women was retrospectively assembled. All mammograms were originally interpreted as negative by radiologists. In the next subsequent mammography screening, 402 women were diagnosed with breast cancer and 642 remained negative. An existing CAD scheme was applied 'as is' to process each image. From CAD-generated results, four detection features including the total number of (1) initial detection seeds and (2) the final detected false-positive regions, (3) average and (4) sum of detection scores, were computed from each image. Then, by combining the features computed from two bilateral images of left and right breasts from either craniocaudal or mediolateral oblique view, two logistic regression models were trained and tested using a leave-one-case-out cross-validation method to predict the likelihood of each testing case being positive in the next subsequent screening. The new prediction model yielded the maximum prediction accuracy with an area under a ROC curve of AUC  =  0.65  ±  0.017 and the maximum adjusted odds ratio of 4.49 with a 95% confidence interval of (2.95, 6.83). The results also showed an increasing trend in the adjusted odds ratio and risk prediction scores (p  <  0.01). Thus, this study demonstrated that CAD-generated false-positives might include valuable information, which needs to be further explored for identifying and/or developing more effective imaging markers for predicting short-term breast cancer risk.

Mammographic breast density in recent and longer-standing ethiopian immigrants to israel

High breast density is associated with an increased risk of breast cancer development. Little is known concerning ethnic variations in breast density and its relevant contributing factors. We aimed to study breast density among Ethiopian immigrants to Israel in comparison with Israeli-born women and to determine any effect on breast density of the length of residency in the immigrant population. Mammographic breast density using the BI-RADS system was estimated and compared between 77 women of Ethiopian origin who live in Israel and 177 Israeli-born controls. Logistic regression analysis was performed to estimate the odds ratios (OR) for high density (BI-RADS score ≥ 3) vs low density (BI-RADS score < 3) cases, comparing the 2 origin groups. Ethiopian-born women had a crude OR of 0.15 (95% CI: 0.08-0.26) for high breast density compared with Israeli-born women. Adjustments for various cofounders did not affect the results. Time since immigration to Israel seemed to modify the relationship, with a stronger association for women who immigrated within 2 years prior to mammography (OR:0.07, 95% CI: 0.03-0.17) as opposed to women with a longer residency stay in Israel (OR:0.23, 95% CI:0.10-0.50). Adjustments of various confounders did not alter these findings. Breast density in Ethiopian immigrants to Israel is significantly lower than that of Israeli-born controls. Our study suggests a positive association between time since immigration and breast density. Future studies are required to define the possible effects of dietary change on mammographic density following immigration.

A new near-term breast cancer risk prediction scheme based on the quantitative analysis of ipsilateral view mammograms

PURPOSE

To help improve efficacy of screening mammography and eventually establish an optimal personalized screening paradigm, this study aimed to develop and test a new near-term breast cancer risk prediction scheme based on the quantitative analysis of ipsilateral view of the negative screening mammograms.

METHODS

The dataset includes digital mammograms acquired from 392 women with two sequential full-field digital mammography examinations. All the first ("prior") sets of mammograms were interpreted as negative during the original reading. In the sequential ("current") screening, 202 were proved positive and 190 remained negative/benign. For each pair of the "prior" ipsilateral mammograms, we adaptively fused the image features computed from two views. Using four different types of image features, we built four elastic net support vector machine (EnSVM) based classifiers. Then, the initial prediction scores form the 4 EnSVMs were combined to build a final artificial neural network (ANN) classifier that produces the final risk prediction score. The performance of the new scheme was evaluated by using a 10-fold cross-validation method and an assessment index of the area under the receiver operating characteristic curve (AUC).

RESULTS

A total number of 466 features were initially extracted from each pair of ipsilateral mammograms. Among them, 51 were selected to build the EnSVM based prediction scheme. The AUC = 0.737 ± 0.052 was yielded using the new scheme. Applying an optimal operating threshold, the prediction sensitivity was 60.4% (122 of 202) and the specificity was 79.0% (150 of 190).

CONCLUSION

The study results showed moderately high positive association between computed risk scores using the "prior" negative mammograms and the actual outcome of the image-detectable breast cancers in the next subsequent screening examinations. The study also demonstrated that quantitative analysis of the ipsilateral views of the mammograms enabled to provide useful information in predicting near-term breast cancer risk.

Prediction of breast cancer risk using a machine learning approach embedded with a locality preserving projection algorithm

In order to automatically identify a set of effective mammographic image features and build an optimal breast cancer risk stratification model, this study aims to investigate advantages of applying a machine learning approach embedded with a locally preserving projection (LPP) based feature combination and regeneration algorithm to predict short-term breast cancer risk. A dataset involving negative mammograms acquired from 500 women was assembled. This dataset was divided into two age-matched classes of 250 high risk cases in which cancer was detected in the next subsequent mammography screening and 250 low risk cases, which remained negative. First, a computer-aided image processing scheme was applied to segment fibro-glandular tissue depicted on mammograms and initially compute 44 features related to the bilateral asymmetry of mammographic tissue density distribution between left and right breasts. Next, a multi-feature fusion based machine learning classifier was built to predict the risk of cancer detection in the next mammography screening. A leave-one-case-out (LOCO) cross-validation method was applied to train and test the machine learning classifier embedded with a LLP algorithm, which generated a new operational vector with 4 features using a maximal variance approach in each LOCO process. Results showed a 9.7% increase in risk prediction accuracy when using this LPP-embedded machine learning approach. An increased trend of adjusted odds ratios was also detected in which odds ratios increased from 1.0 to 11.2. This study demonstrated that applying the LPP algorithm effectively reduced feature dimensionality, and yielded higher and potentially more robust performance in predicting short-term breast cancer risk.

Deep learning in breast cancer risk assessment: evaluation of convolutional neural networks on a clinical dataset of full-field digital mammograms

To evaluate deep learning in the assessment of breast cancer risk in which convolutional neural networks (CNNs) with transfer learning are used to extract parenchymal characteristics directly from full-field digital mammographic (FFDM) images instead of using computerized radiographic texture analysis (RTA), 456 clinical FFDM cases were included: a "high-risk" BRCA1/2 gene-mutation carriers dataset (53 cases), a "high-risk" unilateral cancer patients dataset (75 cases), and a "low-risk dataset" (328 cases). Deep learning was compared to the use of features from RTA, as well as to a combination of both in the task of distinguishing between high- and low-risk subjects. Similar classification performances were obtained using CNN [area under the curve [Formula: see text]; standard error [Formula: see text]] and RTA ([Formula: see text]; [Formula: see text]) in distinguishing BRCA1/2 carriers and low-risk women. However, in distinguishing unilateral cancer patients and low-risk women, performance was significantly greater with CNN ([Formula: see text]; [Formula: see text]) compared to RTA ([Formula: see text]; [Formula: see text]). Fusion classifiers performed significantly better than the RTA-alone classifiers with AUC values of 0.86 and 0.84 in differentiating BRCA1/2 carriers from low-risk women and unilateral cancer patients from low-risk women, respectively. In conclusion, deep learning extracted parenchymal characteristics from FFDMs performed as well as, or better than, conventional texture analysis in the task of distinguishing between cancer risk populations.

Breast Tissue Organisation and its Association with Breast Cancer Risk

Background

Mammographic percentage density is an established and important risk factor for breast cancer. In this paper, we investigate the role of the spatial organisation of (dense vs. fatty) regions of the breast defined from mammographic images in terms of breast cancer risk.

Methods

We present a novel approach that provides a thorough description of the spatial organisation of different types of tissue in the breast. Each mammogram is first segmented into four regions (fatty, semi-fatty, semi-dense and dense tissue). The spatial relations between each pair of regions is described using so-called forces histograms (FHs) and summarised using functional principal component analysis. In our main analysis, association with case–control status is assessed using a Swedish population-based case–control study (1,170 cases and 1283 controls), for which digitised mammograms were available. We also carried out a small validation study based on digital images.

Results

For our main analysis, we obtained a global p value of 2×10⁻⁷ indicating a significant association between the spatial relations of the four segmented regions and breast cancer status after adjustment for percentage density and other important breast cancer risk factors. Our (spatial relations) score had a per standard deviation odds ratio 1.29, after accounting for overfitting (percentage density had a per standard deviation odds ratio of 1.34). The spatial relations between the fatty and semi-fatty tissue and the spatial relations between the fatty and dense tissue were the most significant. The spatial relations between the fatty and semi-fatty tissue were associated with parity and age at first birth (p=6×10⁻⁴). Using digital images, we were able to verify that the same characteristics of tissue organisation can be identified and we validated the association for the spatial relations between the fatty and semi-fatty tissue.

Conclusions

Our findings are consistent with the notion that fibroglandular and adipose tissue plays a role in breast cancer risk and, more specifically, they suggest that fatty tissue in the lower quadrants and the absence of density in the retromammary space, as shown in mediolateral oblique images, are protective against breast cancer.

Breast density estimation from high spectral and spatial resolution MRI

A three-dimensional breast density estimation method is presented for high spectral and spatial resolution (HiSS) MR imaging. Twenty-two patients were recruited (under an Institutional Review Board--approved Health Insurance Portability and Accountability Act-compliant protocol) for high-risk breast cancer screening. Each patient received standard-of-care clinical digital x-ray mammograms and MR scans, as well as HiSS scans. The algorithm for breast density estimation includes breast mask generating, breast skin removal, and breast percentage density calculation. The inter- and intra-user variabilities of the HiSS-based density estimation were determined using correlation analysis and limits of agreement. Correlation analysis was also performed between the HiSS-based density estimation and radiologists' breast imaging-reporting and data system (BI-RADS) density ratings. A correlation coefficient of 0.91 ([Formula: see text]) was obtained between left and right breast density estimations. An interclass correlation coefficient of 0.99 ([Formula: see text]) indicated high reliability for the inter-user variability of the HiSS-based breast density estimations. A moderate correlation coefficient of 0.55 ([Formula: see text]) was observed between HiSS-based breast density estimations and radiologists' BI-RADS. In summary, an objective density estimation method using HiSS spectral data from breast MRI was developed. The high reproducibility with low inter- and low intra-user variabilities shown in this preliminary study suggest that such a HiSS-based density metric may be potentially beneficial in programs requiring breast density such as in breast cancer risk assessment and monitoring effects of therapy.

Mammographic texture and risk of breast cancer by tumor type and estrogen receptor status

Background

Several studies have shown that mammographic texture features are associated with breast cancer risk independent of the contribution of breast density. Thus, texture features may provide novel information for risk stratification. We examined the association of a set of established texture features with breast cancer risk by tumor type and estrogen receptor (ER) status, accounting for breast density.

Methods

This study combines five case–control studies including 1171 breast cancer cases and 1659 controls matched for age, date of mammogram, and study. Mammographic breast density and 46 breast texture features, including first- and second-order features, Fourier transform, and fractal dimension analysis, were evaluated from digitized film-screen mammograms. Logistic regression models evaluated each normalized feature with breast cancer after adjustment for age, body mass index, first-degree family history, percent density, and study.

Results

Of the mammographic features analyzed, fractal dimension and second-order statistics features were significantly associated (p < 0.05) with breast cancer. Fractal dimensions for the thresholds equal to 10% and 15% (FD_TH10 and FD_TH15) were associated with an increased risk of breast cancer while thresholds from 60% to 85% (FD_TH60 to FD_TH85) were associated with a decreased risk. Increasing the FD_TH75 and Energy feature values were associated with a decreased risk of breast cancer while increasing Entropy was associated with a decreased risk of breast cancer. For example, 1 standard deviation increase of FD_TH75 was associated with a 13% reduced risk of breast cancer (odds ratio = 0.87, 95% confidence interval 0.79–0.95). Overall, the direction of associations between features and ductal carcinoma in situ (DCIS) and invasive cancer, and estrogen receptor positive and negative cancer were similar.

Conclusion

Mammographic features derived from film-screen mammograms are associated with breast cancer risk independent of percent mammographic density. Some texture features also demonstrated associations for specific tumor types. For future work, we plan to assess risk prediction combining mammographic density and features assessed on digital images.

Electronic supplementary material

The online version of this article (doi:10.1186/s13058-016-0778-1) contains supplementary material, which is available to authorized users.

Beyond breast density: a review on the advancing role of parenchymal texture analysis in breast cancer risk assessment

BACKGROUND

The assessment of a woman's risk for developing breast cancer has become increasingly important for establishing personalized screening recommendations and forming preventive strategies. Studies have consistently shown a strong relationship between breast cancer risk and mammographic parenchymal patterns, typically assessed by percent mammographic density. This paper will review the advancing role of mammographic texture analysis as a potential novel approach to characterize the breast parenchymal tissue to augment conventional density assessment in breast cancer risk estimation.

MAIN TEXT

The analysis of mammographic texture provides refined, localized descriptors of parenchymal tissue complexity. Currently, there is growing evidence in support of textural features having the potential to augment the typically dichotomized descriptors (dense or not dense) of area or volumetric measures of breast density in breast cancer risk assessment. Therefore, a substantial research effort has been devoted to automate mammographic texture analysis, with the aim of ultimately incorporating such quantitative measures into breast cancer risk assessment models. In this paper, we review current and emerging approaches in this field, summarizing key methodological details and related studies using novel computerized approaches. We also discuss research challenges for advancing the role of parenchymal texture analysis in breast cancer risk stratification and accelerating its clinical translation.

CONCLUSIONS

The objective is to provide a comprehensive reference for researchers in the field of parenchymal pattern analysis in breast cancer risk assessment, while indicating key directions for future research.

Using multiscale texture and density features for near-term breast cancer risk analysis

PURPOSE

To help improve efficacy of screening mammography by eventually establishing a new optimal personalized screening paradigm, the authors investigated the potential of using the quantitative multiscale texture and density feature analysis of digital mammograms to predict near-term breast cancer risk.

METHODS

The authors' dataset includes digital mammograms acquired from 340 women. Among them, 141 were positive and 199 were negative/benign cases. The negative digital mammograms acquired from the "prior" screening examinations were used in the study. Based on the intensity value distributions, five subregions at different scales were extracted from each mammogram. Five groups of features, including density and texture features, were developed and calculated on every one of the subregions. Sequential forward floating selection was used to search for the effective combinations. Using the selected features, a support vector machine (SVM) was optimized using a tenfold validation method to predict the risk of each woman having image-detectable cancer in the next sequential mammography screening. The area under the receiver operating characteristic curve (AUC) was used as the performance assessment index.

RESULTS

From a total number of 765 features computed from multiscale subregions, an optimal feature set of 12 features was selected. Applying this feature set, a SVM classifier yielded performance of AUC = 0.729 ± 0.021. The positive predictive value was 0.657 (92 of 140) and the negative predictive value was 0.755 (151 of 200).

CONCLUSIONS

The study results demonstrated a moderately high positive association between risk prediction scores generated by the quantitative multiscale mammographic image feature analysis and the actual risk of a woman having an image-detectable breast cancer in the next subsequent examinations.

Parenchymal texture analysis in digital mammography: A fully automated pipeline for breast cancer risk assessment

PURPOSE

Mammographic percent density (PD%) is known to be a strong risk factor for breast cancer. Recent studies also suggest that parenchymal texture features, which are more granular descriptors of the parenchymal pattern, can provide additional information about breast cancer risk. To date, most studies have measured mammographic texture within selected regions of interest (ROIs) in the breast, which cannot adequately capture the complexity of the parenchymal pattern throughout the whole breast. To better characterize patterns of the parenchymal tissue, the authors have developed a fully automated software pipeline based on a novel lattice-based strategy to extract a range of parenchymal texture features from the entire breast region.

METHODS

Digital mammograms from 106 cases with 318 age-matched controls were retrospectively analyzed. The lattice-based approach is based on a regular grid virtually overlaid on each mammographic image. Texture features are computed from the intersection (i.e., lattice) points of the grid lines within the breast, using a local window centered at each lattice point. Using this strategy, a range of statistical (gray-level histogram, co-occurrence, and run-length) and structural (edge-enhancing, local binary pattern, and fractal dimension) features are extracted. To cover the entire breast, the size of the local window for feature extraction is set equal to the lattice grid spacing and optimized experimentally by evaluating different windows sizes. The association between their lattice-based texture features and breast cancer was evaluated using logistic regression with leave-one-out cross validation and further compared to that of breast PD% and commonly used single-ROI texture features extracted from the retroareolar or the central breast region. Classification performance was evaluated using the area under the curve (AUC) of the receiver operating characteristic (ROC). DeLong's test was used to compare the different ROCs in terms of AUC performance.

RESULTS

The average univariate performance of the lattice-based features is higher when extracted from smaller than larger window sizes. While not every individual texture feature is superior to breast PD% (AUC: 0.59, STD: 0.03), their combination in multivariate analysis has significantly better performance (AUC: 0.85, STD: 0.02, p < 0.001). The lattice-based texture features also outperform the single-ROI texture features when extracted from the retroareolar or the central breast region (AUC: 0.60-0.74, STD: 0.03). Adding breast PD% does not make a significant performance improvement to the lattice-based texture features or the single-ROI features (p > 0.05).

CONCLUSIONS

The proposed lattice-based strategy for mammographic texture analysis enables to characterize the parenchymal pattern over the entire breast. As such, these features provide richer information compared to currently used descriptors and may ultimately improve breast cancer risk assessment. Larger studies are warranted to validate these findings and also compare to standard demographic and reproductive risk factors.

A new approach to develop computer-aided detection schemes of digital mammograms

The purpose of this study is to develop a new global mammographic image feature analysis based computer-aided detection (CAD) scheme and evaluate its performance in detecting positive screening mammography examinations. A dataset that includes images acquired from 1896 full-field digital mammography (FFDM) screening examinations was used in this study. Among them, 812 cases were positive for cancer and 1084 were negative or benign. After segmenting the breast area, a computerized scheme was applied to compute 92 global mammographic tissue density based features on each of four mammograms of the craniocaudal (CC) and mediolateral oblique (MLO) views. After adding three existing popular risk factors (woman's age, subjectively rated mammographic density, and family breast cancer history) into the initial feature pool, we applied a sequential forward floating selection feature selection algorithm to select relevant features from the bilateral CC and MLO view images separately. The selected CC and MLO view image features were used to train two artificial neural networks (ANNs). The results were then fused by a third ANN to build a two-stage classifier to predict the likelihood of the FFDM screening examination being positive. CAD performance was tested using a ten-fold cross-validation method. The computed area under the receiver operating characteristic curve was AUC = 0.779   ±   0.025 and the odds ratio monotonically increased from 1 to 31.55 as CAD-generated detection scores increased. The study demonstrated that this new global image feature based CAD scheme had a relatively higher discriminatory power to cue the FFDM examinations with high risk of being positive, which may provide a new CAD-cueing method to assist radiologists in reading and interpreting screening mammograms.

Assessment of a Four-View Mammographic Image Feature Based Fusion Model to Predict Near-Term Breast Cancer Risk

The purpose of this study was to develop and assess a new quantitative four-view mammographic image feature based fusion model to predict the near-term breast cancer risk of the individual women after a negative screening mammography examination of interest. The dataset included fully-anonymized mammograms acquired on 870 women with two sequential full-field digital mammography examinations. For each woman, the first "prior" examination in the series was interpreted as negative (not recalled) during the original image reading. In the second "current" examination, 430 women were diagnosed with pathology verified cancers and 440 remained negative ("cancer-free"). For each of four bilateral craniocaudal and mediolateral oblique view images of left and right breasts, we computed and analyzed eight groups of global mammographic texture and tissue density image features. A risk prediction model based on three artificial neural networks was developed to fuse image features computed from two bilateral views of four images. The risk model performance was tested using a ten-fold cross-validation method and a number of performance evaluation indices including the area under the receiver operating characteristic curve (AUC) and odds ratio (OR). The highest AUC = 0.725 ± 0.026 was obtained when the model was trained by gray-level run length statistics texture features computed on dense breast regions, which was significantly higher than the AUC values achieved using the model trained by only two bilateral one-view images (p < 0.02). The adjustable OR values monotonically increased from 1.0 to 11.8 as model-generated risk score increased. The regression analysis of OR values also showed a significant increase trend in slope (p < 0.01). As a result, this preliminary study demonstrated that a new four-view mammographic image feature based risk model could provide useful and supplementary image information to help predict the near-term breast cancer risk.

Comparative analysis of image-based phenotypes of mammographic density and parenchymal patterns in distinguishing between BRCA1/2 cases, unilateral cancer cases, and controls

We statistically compare the contributions of parenchymal phenotypes to mammographic density in distinguishing between high-risk cases and low-risk controls. The age-matched evaluation included computerized mammographic assessment of breast percent density (PD) and parenchymal patterns (phenotypes of coarseness and contrast) from radiographic texture analysis (RTA) of the full-field digital mammograms from 456 cases: 53 women with BRCA1/2 gene mutations, 75 with unilateral cancer, and 328 at low risk of developing breast cancer. Image-based phenotypes of parenchymal pattern coarseness and contrast were each found to significantly discriminate between the groups; however, PD did not. From ROC analysis, PD alone yielded area under the fitted ROC curve (AUC) values of 0.53 ([Formula: see text]) and 0.57 ([Formula: see text]) in the classification task between BRCA1/2 gene-mutation carriers and low-risk women, and between unilateral cancer and low-risk women, respectively. In a round-robin evaluation with Bayesian artificial neural network (BANN) analysis, RTA yielded AUC values of 0.81 (95% confidence interval [0.71, 0.89]) and 0.70 (95% confidence interval [0.63, 0.77]) between the BRCA1/2 gene-mutation carriers and low-risk women, and between unilateral cancer and low-risk women, respectively. These results show that high-risk and low-risk women have different mammographic parenchymal patterns with significantly higher discrimination resulting from characteristics of the parenchymal patterns than just the breast PD.

A method to determine the mammographic regions that show early changes due to the development of breast cancer

It is well understood nowadays that changes in the mammographic parenchymal pattern are an indicator of a risk of breast cancer and we have developed a statistical method that estimates the mammogram regions where the parenchymal changes, due to breast cancer, occur. This region of interest is computed from a score map by utilising the anatomical breast coordinate system developed in our previous work. The method also makes an automatic scale selection to avoid overfitting while the region estimates are computed by a nested cross-validation scheme. In this way, it is possible to recover those mammogram regions that show a significant difference in classification scores between the cancer and the control group. Our experiments suggested that the most significant mammogram region is the region behind the nipple and that can be justified by previous findings from other research groups. This result was conducted on the basis of the cross-validation experiments on independent training, validation and testing sets from the case-control study of 490 women, of which 245 women were diagnosed with breast cancer within a period of 2-4 years after the baseline mammograms. We additionally generalised the estimated region to another, mini-MIAS study and showed that the transferred region estimate gives at least a similar classification result when compared to the case where the whole breast region is used. In all, by following our method, one most likely improves both preclinical and follow-up breast cancer screening, but a larger study population will be required to test this hypothesis.

Relationships between computer-extracted mammographic texture pattern features and BRCA1/2 mutation status: a cross-sectional study

INTRODUCTION

Mammographic density is similar among women at risk of either sporadic or BRCA1/2-related breast cancer. It has been suggested that digitized mammographic images contain computer-extractable information within the parenchymal pattern, which may contribute to distinguishing between BRCA1/2 mutation carriers and non-carriers.

METHODS

We compared mammographic texture pattern features in digitized mammograms from women with deleterious BRCA1/2 mutations (n = 137) versus non-carriers (n = 100). Subjects were stratified into training (107 carriers, 70 non-carriers) and testing (30 carriers, 30 non-carriers) datasets. Masked to mutation status, texture features were extracted from a retro-areolar region-of-interest in each subject's digitized mammogram. Stepwise linear regression analysis of the training dataset identified variables to be included in a radiographic texture analysis (RTA) classifier model aimed at distinguishing BRCA1/2 carriers from non-carriers. The selected features were combined using a Bayesian Artificial Neural Network (BANN) algorithm, which produced a probability score rating the likelihood of each subject's belonging to the mutation-positive group. These probability scores were evaluated in the independent testing dataset to determine whether their distribution differed between BRCA1/2 mutation carriers and non-carriers. A receiver operating characteristic analysis was performed to estimate the model's discriminatory capacity.

RESULTS

In the testing dataset, a one standard deviation (SD) increase in the probability score from the BANN-trained classifier was associated with a two-fold increase in the odds of predicting BRCA1/2 mutation status: unadjusted odds ratio (OR) = 2.00, 95% confidence interval (CI): 1.59, 2.51, P = 0.02; age-adjusted OR = 1.93, 95% CI: 1.53, 2.42, P = 0.03. Additional adjustment for percent mammographic density did little to change the OR. The area under the curve for the BANN-trained classifier to distinguish between BRCA1/2 mutation carriers and non-carriers was 0.68 for features alone and 0.72 for the features plus percent mammographic density.

CONCLUSIONS

Our findings suggest that, unlike percent mammographic density, computer-extracted mammographic texture pattern features are associated with carrying BRCA1/2 mutations. Although still at an early stage, our novel RTA classifier has potential for improving mammographic image interpretation by permitting real-time risk stratification among women undergoing screening mammography.

Relationships between computer-extracted mammographic texture pattern features and BRCA1/2 mutation status: a cross-sectional study

Introduction

Mammographic density is similar among women at risk of either sporadic or BRCA1/2-related breast cancer. It has been suggested that digitized mammographic images contain computer-extractable information within the parenchymal pattern, which may contribute to distinguishing between BRCA1/2 mutation carriers and non-carriers.

Methods

We compared mammographic texture pattern features in digitized mammograms from women with deleterious BRCA1/2 mutations (n = 137) versus non-carriers (n = 100). Subjects were stratified into training (107 carriers, 70 non-carriers) and testing (30 carriers, 30 non-carriers) datasets. Masked to mutation status, texture features were extracted from a retro-areolar region-of-interest in each subject’s digitized mammogram. Stepwise linear regression analysis of the training dataset identified variables to be included in a radiographic texture analysis (RTA) classifier model aimed at distinguishing BRCA1/2 carriers from non-carriers. The selected features were combined using a Bayesian Artificial Neural Network (BANN) algorithm, which produced a probability score rating the likelihood of each subject’s belonging to the mutation-positive group. These probability scores were evaluated in the independent testing dataset to determine whether their distribution differed between BRCA1/2 mutation carriers and non-carriers. A receiver operating characteristic analysis was performed to estimate the model’s discriminatory capacity.

Results

In the testing dataset, a one standard deviation (SD) increase in the probability score from the BANN-trained classifier was associated with a two-fold increase in the odds of predicting BRCA1/2 mutation status: unadjusted odds ratio (OR) = 2.00, 95% confidence interval (CI): 1.59, 2.51, P = 0.02; age-adjusted OR = 1.93, 95% CI: 1.53, 2.42, P = 0.03. Additional adjustment for percent mammographic density did little to change the OR. The area under the curve for the BANN-trained classifier to distinguish between BRCA1/2 mutation carriers and non-carriers was 0.68 for features alone and 0.72 for the features plus percent mammographic density.

Conclusions

Our findings suggest that, unlike percent mammographic density, computer-extracted mammographic texture pattern features are associated with carrying BRCA1/2 mutations. Although still at an early stage, our novel RTA classifier has potential for improving mammographic image interpretation by permitting real-time risk stratification among women undergoing screening mammography.

Electronic supplementary material

The online version of this article (doi:10.1186/s13058-014-0424-8) contains supplementary material, which is available to authorized users.

Reduction of false-positive recalls using a computerized mammographic image feature analysis scheme

The high false-positive recall rate is one of the major dilemmas that significantly reduce the efficacy of screening mammography, which harms a large fraction of women and increases healthcare cost. This study aims to investigate the feasibility of helping reduce false-positive recalls by developing a new computer-aided diagnosis (CAD) scheme based on the analysis of global mammographic texture and density features computed from four-view images. Our database includes full-field digital mammography (FFDM) images acquired from 1052 recalled women (669 positive for cancer and 383 benign). Each case has four images: two craniocaudal (CC) and two mediolateral oblique (MLO) views. Our CAD scheme first computed global texture features related to the mammographic density distribution on the segmented breast regions of four images. Second, the computed features were given to two artificial neural network (ANN) classifiers that were separately trained and tested in a ten-fold cross-validation scheme on CC and MLO view images, respectively. Finally, two ANN classification scores were combined using a new adaptive scoring fusion method that automatically determined the optimal weights to assign to both views. CAD performance was tested using the area under a receiver operating characteristic curve (AUC). The AUC = 0.793  ±  0.026 was obtained for this four-view CAD scheme, which was significantly higher at the 5% significance level than the AUCs achieved when using only CC (p = 0.025) or MLO (p = 0.0004) view images, respectively. This study demonstrates that a quantitative assessment of global mammographic image texture and density features could provide useful and/or supplementary information to classify between malignant and benign cases among the recalled cases, which may eventually help reduce the false-positive recall rate in screening mammography.

Segmentation of breast masses on dedicated breast computed tomography and three-dimensional breast ultrasound images

We present and evaluate a method for the three-dimensional (3-D) segmentation of breast masses on dedicated breast computed tomography (bCT) and automated 3-D breast ultrasound images. The segmentation method, refined from our previous segmentation method for masses on contrast-enhanced bCT, includes two steps: (1) initial contour estimation and (2) active contour-based segmentation to further evolve and refine the initial contour by adding a local energy term to the level-set equation. Segmentation performance was assessed in terms of Dice coefficients (DICE) for 129 lesions on noncontrast bCT, 38 lesions on contrast-enhanced bCT, and 98 lesions on 3-D breast ultrasound (US) images. For bCT, DICE values of 0.82 and 0.80 were obtained on contrast-enhanced and noncontrast images, respectively. The improvement in segmentation performance with respect to that of our previous method was statistically significant ( p = 0.002 ). Moreover, segmentation appeared robust with respect to the presence of glandular tissue. For 3-D breast US, the DICE value was 0.71. Hence, our method obtained promising results for both 3-D imaging modalities, laying a solid foundation for further quantitative image analysis and potential future expansion to other 3-D imaging modalities.

Enhancement of mammographic density measures in breast cancer risk prediction

BACKGROUND

Mammographic density is a strong risk factor for breast cancer.

METHODS

We present a novel approach to enhance area density measures that takes advantage of the relative density of the pectoral muscle that appears in lateral mammographic views. We hypothesized that the grey scale of film mammograms is normalized to volume breast density but not pectoral density and thus pectoral density becomes an independent marker of volumetric density.

RESULTS

From analysis of data from a Swedish case-control study (1,286 breast cancer cases and 1,391 control subjects, ages 50-75 years), we found that the mean intensity of the pectoral muscle (MIP) was highly associated with breast cancer risk [per SD: OR = 0.82; 95% confidence interval (CI), 0.75-0.88; P = 6 × 10(-7)] after adjusting for a validated computer-assisted measure of percent density (PD), Cumulus. The area under curve (AUC) changed from 0.600 to 0.618 due to using PD with the pectoral muscle as reference instead of a standard area-based PD measure. We showed that MIP is associated with a genetic variant known to be associated with mammographic density and breast cancer risk, rs10995190, in a subset of women with genetic data. We further replicated the association between MIP and rs10995190 in an additional cohort of 2,655 breast cancer cases (combined P = 0.0002).

CONCLUSIONS

MIP is a marker of volumetric density that can be used to complement area PD in mammographic density studies and breast cancer risk assessment.

IMPACT

Inclusion of MIP in risk models should be considered for studies using area PD from analog films.

Mammographic texture resemblance generalizes as an independent risk factor for breast cancer

Introduction

Breast density has been established as a major risk factor for breast cancer. We have previously demonstrated that mammographic texture resemblance (MTR), recognizing the local texture patterns of the mammogram, is also a risk factor for breast cancer, independent of percent breast density. We examine if these findings generalize to another population.

Methods

Texture patterns were recorded in digitalized pre-diagnosis (3.7 years) film mammograms of a nested case–control study within the Dutch screening program (S1) comprising of 245 breast cancers and 250 matched controls. The patterns were recognized in the same study using cross-validation to form resemblance scores associated with breast cancer. Texture patterns from S1 were examined in an independent nested case–control study within the Mayo Mammography Health Study cohort (S2) of 226 cases and 442 matched controls: mammograms on average 8.5 years prior to diagnosis, risk factor information and percent mammographic density (PD) estimated using Cumulus were available. MTR scores estimated from S1, S2 and S1 + S2 (the latter two as cross-validations) were evaluated in S2. MTR scores were analyzed as both quartiles and continuously for association with breast cancer using odds ratios (OR) and adjusting for known risk factors including age, body mass index (BMI), and hormone usage.

Results

The mean ages of S1 and S2 were 58.0 ± 5.7 years and 55.2 ± 10.5 years, respectively. The MTR scores on S1 showed significant capability to discriminate cancers from controls (area under the operator characteristics curve (AUC) = 0.63 ± 0.02, P <0.001), which persisted after adjustment for PD. S2 showed an AUC of 0.63, 0.61, and 0.60 based on PD, MTR scores trained on S2, and MTR scores trained on S1, respectively. When adjusted for PD, MTR scores of S2 trained on S1 showed an association with breast cancer for the highest quartile alone: OR in quartiles of controls as reference; 1.04 (0.59 to 1.81); 0.95 (0.52 to 1.74); 1.84 (1.10 to 3.07) respectively. The combined continuous model with both PD and MTR scores based on S1 had an AUC of 0.66 ± 0.03.

Conclusions

The local texture patterns associated with breast cancer risk in S1 were also an independent risk factor in S2. Additional textures identified in S2 did not significantly improve risk segregation. Hence, the textural patterns that indicated elevated risk persisted under differences in X-ray technology, population demographics, follow-up time and geography.