Identifying clinically relevant findings in breast cancer using deep learning and feature attribution on local views from high-resolution mammography

Mammography for Older Women?

Purpose: The prevalence of artificial intelligence (AI) in breast cancer screening and diagnosis has substantially increased over the past 5 years. AI can be used to assist in the interpretation of mammograms and other breast imaging modalities, as well as predict patient outcomes. This study aims to quantify AI's diagnostic accuracy and its ability to detect breast cancer from a large database of mammograms. The implementation of AI in breast cancer screening holds immense potential to not only enhance the accuracy of mammogram interpretation but also improve the overall efficiency of radiologists and healthcare providers. Furthermore, we aim to contribute to the advancement of the rapidly growing field of machine learning in medicine, particularly in breast cancer care. Our hope is that by applying this technology, we can improve patient outcomes and increase accessibility to breast cancer screening and early detection. Methods: A deep neural network (DNN) model was built on Tensorflow utilizing Radiological Society of North America Screening Mammography Breast Cancer Detection dataset, consisting of normal mammograms and abnormal mammograms with tumors. The model's efficacy was assessed using Area under the curve (AUC), precision, and recall, with images randomly split into training (80%), validation (10%), and test sets (10%). Four consecutive training sessions each lasting 2 hours and 13 minutes using 8,877 images consisting of 4,621 breast cancer and 4,456 normal breast mammogram images was done. Results: The model attained an AUC of 0.926, Specificity of 95%, Sensitivity of 73% and Accuracy of 84%. Conclusion: This DNN model was created for diagnosing breast cancer from mammograms, with higher AUC than the current standard. This study highlights the potential of AI to revolutionize breast cancer screening, prevention, and diagnosis, which will ultimately improve patient outcomes and increase accessibility for patients. The use of AI in detecting breast cancer on mammograms can not only provide support to radiologists but also improve their efficacy and accuracy, reducing the burden of high volume images. Additionally, the use of AI in detection and screening can address the barriers that come with physician shortages, especially in underserved areas. Extending AI’s reach to these areas of limited access can improve early detection and screening which tends to be lacking in under-resourced areas, thus addressing the gap in health equity worldwide. While this study highlights encouraging data at the intersection of AI and medicine, we acknowledge that AI’s role in medicine is not to replace but rather enhance the expertise of radiologists. Lastly, we hope that our developed tool can alleviate the workload of radiologists and increase efficiency in high-pressure settings such as emergency departments and rural areas, ultimately leading to improved patient care. Citation Format: Parsa Riazi Esfahani, Maya M Maalouf, Akshay J Reddy, Prashant Chawla. Utilizing Machine Learning Techniques to Investigate Mammograms for Breast Cancer Detection [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr A088.

Comprehensive Study for Breast Cancer Using Deep Learning and Traditional Machine Learning

Screening for breast cancer with mammography has been introduced in various countries over the last 30 years, initially using analog screen-film-based systems and, over the last 20 years, transitioning to the use of fully digital systems. With the introduction of digitization, the computer interpretation of images has been a subject of intense interest, resulting in the introduction of computer-aided detection (CADe) and diagnosis (CADx) algorithms in the early 2000's. Although they were introduced with high expectations, the potential improvement in the clinical realm failed to materialize, mostly due to the high number of false positive marks per analyzed image. In the last five years, the artificial intelligence (AI) revolution in computing, driven mostly by deep learning and convolutional neural networks, has also pervaded the field of automated breast cancer detection in digital mammography and digital breast tomosynthesis. Research in this area first involved comparison of its capabilities to that of conventional CADe/CADx methods, which quickly demonstrated the potential of this new technology. In the last couple of years, more mature and some commercial products have been developed, and studies of their performance compared to that of experienced breast radiologists are showing that these algorithms are on par with human-performance levels in retrospective data sets. Although additional studies, especially prospective evaluations performed in the real screening environment, are needed, it is becoming clear that AI will have an important role in the future breast cancer screening realm. Exactly how this new player will shape this field remains to be determined, but recent studies are already evaluating different options for implementation of this technology. The aim of this review is to provide an overview of the basic concepts and developments in the field AI for breast cancer detection in digital mammography and digital breast tomosynthesis. The pitfalls of conventional methods, and how these are, for the most part, avoided by this new technology, will be discussed. Importantly, studies that have evaluated the current capabilities of AI and proposals for how these capabilities should be leveraged in the clinical realm will be reviewed, while the questions that need to be answered before this vision becomes a reality are posed.

The risk factors of breast cancer among women, such as genetic, family history and lifestyle factors, can be divided into high-, intermediate- and average-risk. Determining these risk factors may actually help in preventing breast cancer occurrence. Besides that, screening of breast cancer which include mammography, can be done in promoting early breast cancer detection. Breast magnetic resonance imaging (MRI) has been recommended as a supplemental screening tool in high risk women. The aim of this study was to identify the significant risk factor of breast cancer among women and also to determine the usefulness of breast MRI as an addition to mammography in detection of breast cancer in high risk women. This retrospective cohort study design was conducted using patients’ data taken from those who underwent mammography for screening or diagnostic purposes in Advanced Medical and Dental Institute, Universiti Sains Malaysia, from 2007 until 2015. Data from 289 subjects were successfully retrieved and analysed based on their risk factors of breast cancer. Meanwhile, data from 120 subjects who had high risks and underwent both mammography and breast MRI were further analysed. There were two significant risk factors of breast cancer seen among the study population: family history of breast cancer (p-value=0.012) and previous history of breast or ovarian cancer (p-value <0.001). Breast MRI demonstrated high sensitivity (90%) while mammography demonstrated high specificity (80%) in detection of breast cancer in all 120 subjects. The number of cases of breast cancer detection using breast MRI [46 (38.3%)] was higher compared to mammography [24 (20.0%)]. However, breast MRI was found to be non-significant as an adjunct tool to mammography in detecting breast cancer in high risk women (p-value=0.189). A comprehensive screening guideline and surveillance of women at high risk is indeed useful and should be implemented to increase cancer detection rate at early stage

Background: In Puerto Rico (PR), breast cancer was reported to be the leading cancer among women, accounting for approximately 31% of new cancer cases and 52% of cancer deaths. Mammography screening can reduce mortality approximately 20% by detecting breast cancer at early stages, for which less aggressive treatments and therapies are available. There is little published data regarding factors in PR that could explain mammography non-adherence and if those factors are different between previously screened and never screened women. Objectives: To analyze factors associated with breast cancer screening non-adherence between never screened and previously screened, but not currently adherent women. Methods: Cultivando La Salud is a breast and cervical cancer screening promotion program originally designed for low-income Hispanic women over 50 years old that was minimally adapted for Hispanic woman over 21 years old living in PR. The main objectives of Cultivando La Salud PR were to: (1) evaluate the effectiveness of Cultivando La Salud PR, an evidence-based educational intervention aimed to increase breast and cervical cancer screening tests; (2) increase breast and cervical cancer screening tests among women participants in the study; (3) increase the capacity of a community-based organization in the implementation and evaluation of evidence-based interventions in cancer prevention. Women were eligible to participate in Cultivando La Salud PR if they were non-adherent to breast cancer screening (women age 40 years or older that reported not having a mammography in the past year) or non-adherent to cervical cancer screening (women age 21 years or older that reported not having a Pap test in the last three years) and in good health. This program took place in Canóvanas, a municipality of PR, during the 2012-2013 year period. For purpose of this analysis, only data from not-adherent women to breast cancer screening guidelines (n=301) were used. Statistical Analysis: Descriptive statistics were performed to characterize the study population. Bivariate analysis was used to assess breast cancer screening non-adherence (never screened vs. previously screened women) as the dependent variable and variables who achieved statistical significance (p <0.05) in the bivariate analysis were then included in an age-adjusted logistic regression model. Age-adjusted Prevalence Odds Ratio (POR) with their 95% confidence intervals were also calculated. Results: Data from the baseline survey indicated that 22.0% of the participants older than 40 years old never had a mammography. Never screened women were 4 times more likely to report having Mi Salud, a governmental healthcare plan (age-adjusted POR=4.1, 95%CI:1.3-7.5) and were 3 times more likely to have a family income of less than $15,000 than women previously screened but not adherent (age-adjusted POR=3.1, 95%CI:1.3-7.5). Never screened women were also approximately 3 times more likely to report not having an usual place for receiving routine health care (age-adjusted POR=2.9, 95%CI:1.5-5.3) and were 2 times more likely to have not have a Pap test in the last three years (age-adjusted POR=2.3, 95%CI:1.2-4.3). Discussion: Never screened women reported significant socioeconomic disparities that might be affecting mammography screening practices. A better understanding of the barriers that prevent breast cancer screening in this group will help in the design of educational interventions and public health policies targeted to increase mammography rates in PR. Citation Format: Aleli M. Ayala-Marin, Vivian Colon-Lopez, Camille Velez, Natalie Fernandez-Espada, Maria E. Fernandez. Never screened: Understanding breast cancer nonadherence in Puerto Rico. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B70. doi:10.1158/1538-7755.DISP13-B70

A vision for improved cancer screening in Nigeria.

Advances in breast cancer detection: S Brünner, B Langfeldt. Springer Verlag, Berlin, Heidelberg, New York, 1990

BACKGROUND AND OBJECTIVES:Despite its relatively low incidence in Saudi Arabia, breast cancer has been the most common cancer among Saudi females for the past 12 consecutive years. The objective of this study was to report the results of the first national public breast cancer screening program in Saudi Arabia.METHODS:Women 40 years of age or older underwent breast cancer screening. Mammograms were scored using the Breast Imaging-Reporting and Data System (BI-RADS). Correlations between imaging findings, risk factors and pathological findings were analyzed.RESULTS:Between September 2007 and April 2008, 1215 women were enrolled. The median age was 45 years, and median body mass index was 31.6 kg/m2. Sixteen cases of cancer were diagnosed. No cancer was diagnosed in 942 women with R1/R2 scores, and only 1 case of cancer was diagnosed in 228 women with R0/R3 scores. However, among 26 women with R4/R5 scores, 50% had malignant disease and 35% had benign lesions. No correlation was found between known risk factors and imaging score or cancer diagnosis.CONCLUSIONS:Public acceptance of the breast cancer screening program was encouraging. Longitudinal follow-up will help in better determining the risk factors relevant to our patient population.

Background:Breast cancer (BC) is the second most common cancer among women. BC screening is usually based on mammography interpreted by radiologists. Recently, some researchers have used deep learning to automatically diagnose BC in mammography and so assist radiologists. The progress of BC detection algorithms can be measured by their performance on public datasets. The CBIS-DDSM is a widely used public dataset composed of scanned mammographies, equally divided into malignant and non-malignant (benign) images. Each image is accompanied by the segmentation of the lesion. Shen et al. (Nature Sci. Rep., 2019) presented a BC detection algorithm using an “end-to-end” approach to train deep neural networks. In this algorithm, a patch classifier is first trained to classify local image patches. The patch classifier's weights are then used to initialize the whole image classifier, that is refined using datasets with the cancer status of the whole image. They achieved an AUC of 0.87 [0.84, 0.90] in classifying CBIS-DDSM images, using their best single-model, single-view breast classifier. They used ResNet (He et al., CVPR 2016) as the basis of their algorithm. Our hypothesis was that replacing the old ResNet with the modern EfficientNet (Tan et al., arXiv 2019) and MobileNetV2 (Sandler et al.,CVPR 2018) would result in greater accuracy. Methods:We tested many different models, to conclude that the best model is obtained using EfficientNet-B4 as the base model, with a MobileNetV2 block at the top, followed by a dense layer with two output categories. We trained the patch classifier using 52,528 patches with 224x224 pixels extracted from CBIS-DDSM. From each image, we extracted 20 patches: 10 patches containing the lesion and 10 from the background (without lesion). The patch classifier weights were then used to initialize the whole image classifier, that was trained using the end-to-end approach with CBIS-DDSM images resized to 1152x896 pixels, with data augmentation. The training was performed using a step learning rate of 1e-4 for the first 20 epochs then 1e-5 for the remaining 10 and batch size of 4, using 10-fold cross-validation. We used 81% of the dataset for training, 9% for validation and 10% for testing. Results:We obtained an AUC of 0.8963±0.06, using a single-model, single-view classifier and without test-time data augmentation. Conclusions:Using EfficientNet and MobileNetV2 as the basis of the BC detection algorithm (instead of ResNet), we obtained an improvement in classifying CBIS-DDSM images into malignant/non-malignant: AUC has increased from 0.87 to 0.896. Our AUC is also larger than other recent papers in the literature, such as Shu et al. (IEEE Trans Med. Image, 2020) that achieved an AUC of 0.838 in the same CBIS-DDSM dataset. Citation Format: Daniel G. Petrini, Carlos Shimizu, Gabriel V. Valente, Guilherme Folgueira, Guilherme A. Novaes, Maria L. Katayama, Pedro Serio, Rosimeire A. Roela, Tatiana C. Tucunduva, Maria Aparecida A. Folgueira, Hae Y. Kim. High-accuracy breast cancer detection in mammography using EfficientNet and end-to-end training [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 181.

This meta-analysis aimed to evaluate the combined effectiveness of Magnetic Resonance Imaging (MRI) and mammography in detecting breast cancer in women with dense breasts. A comprehensive search was conducted across PubMed, Web of Science, and EMBASE databases up to December 31, 2023, to identify relevant studies. Studies focusing on breast cancer detection in women with dense breast tissue and providing data on the sensitivity, specificity, or positive predictive value of combined MRI and mammography screening, or the use of MRI following a negative mammogram, were included. The meta-analysis was conducted using Stata 15.0, and study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool. Ten studies, involving 51,602 participants, were included in the meta-analysis. The combined use of MRI and mammography for breast cancer detection in women with dense breasts yielded a pooled sensitivity of 0.87 (95% CI: 0.79-0.92), specificity of 0.95 (95% CI: 0.89-0.97), positive likelihood ratio of 2.55 (95% CI: 1.45-4.46), negative likelihood ratio of 0.11 (95% CI: 0.07-0.17), diagnostic score of 3.18 (95% CI: 2.35-4.02), and diagnostic ratio of 24.14 (95% CI: 10.44-55.81), and an area under the Summary Receiver Operating Characteristic curve of 0.97 (95% CI: 0.95-0.98). This meta-analysis demonstrated that the combination of MRI and mammography enhanced breast cancer detection in women with dense breasts. This synergistic approach significantly improves detection sensitivity in this high-risk group.

What effect does mammographic breast density have on lesion detection in digital mammography?

Socioeconomic inequalities in breast and thyroid cancer screening in Korea: A nationwide cross-sectional study

C72: Changing patterns of socioeconomic inequalities in women cancer screening in South Korea with ten years follow-up of nationwide cross-sectional study

Breast cancer is the most commonly diagnosed invasive cancer among women both globally and within the United States and the number one cause of cancer-related deaths among women globally ([1, 2][1]). Less than 1% of diagnosed breast cancers occur in men ([2][2]) and, therefore, male breast cancer is