Development and Validation of an AI-driven Mammographic Breast Density Classification Tool Based on Radiologist Consensus

Comparative evaluation of image-based vs. text-based vs. multimodal AI approaches for automatic breast density assessment in mammograms

BACKGROUND AND OBJECTIVES

In the last decade, there has been a growing interest in applying artificial intelligence (AI) systems to breast cancer assessment, including breast density evaluation. However, few models have been developed to integrate textual mammographic reports and mammographic images. Our aims are (1) to generate a natural language processing (NLP)-based AI system, (2) to evaluate an external image-based software, and (3) to develop a multimodal system, using the late fusion approach, by integrating image and text inferences for the automatic classification of breast density according to the American College of Radiology (ACR) guidelines in mammograms and radiological reports.

METHODS

We first compared different NLP models, three based on n-gram term frequency - inverse document frequency and two transformer-based architectures, using 1533 unstructured mammogram reports as a training set and 303 reports as a test set. Subsequently, we evaluated an external image-based software using 303 mammogram images. Finally, we assessed our multimodal system taking into account both text and mammogram images.

RESULTS

Our best NLP model achieved 88 % accuracy, while the external software and the multimodal system achieved 75 % and 80 % accuracy, respectively, in classifying ACR breast densities.

CONCLUSION

Although our multimodal system outperforms the image-based tool, it currently does not improve the results offered by the NLP model for ACR breast density classification. Nevertheless, the promising results observed here open the possibility to more comprehensive studies regarding the utilization of multimodal tools in the assessment of breast density.

Deep-learning model for background parenchymal enhancement classification in contrast-enhanced mammography

Background.Breast background parenchymal enhancement (BPE) is correlated with the risk of breast cancer. BPE level is currently assessed by radiologists in contrast-enhanced mammography (CEM) using 4 classes: minimal, mild, moderate and marked, as described inbreast imaging reporting and data system(BI-RADS). However, BPE classification remains subject to intra- and inter-reader variability. Fully automated methods to assess BPE level have already been developed in breast contrast-enhanced MRI (CE-MRI) and have been shown to provide accurate and repeatable BPE level classification. However, to our knowledge, no BPE level classification tool is available in the literature for CEM.Materials and methods.A BPE level classification tool based on deep learning has been trained and optimized on 7012 CEM image pairs (low-energy and recombined images) and evaluated on a dataset of 1013 image pairs. The impact of image resolution, backbone architecture and loss function were analyzed, as well as the influence of lesion presence and type on BPE assessment. The evaluation of the model performance was conducted using different metrics including 4-class balanced accuracy and mean absolute error. The results of the optimized model for a binary classification: minimal/mild versus moderate/marked, were also investigated.Results.The optimized model achieved a 4-class balanced accuracy of 71.5% (95% CI: 71.2-71.9) with 98.8% of classification errors between adjacent classes. For binary classification, the accuracy reached 93.0%. A slight decrease in model accuracy is observed in the presence of lesions, but it is not statistically significant, suggesting that our model is robust to the presence of lesions in the image for a classification task. Visual assessment also confirms that the model is more affected by non-mass enhancements than by mass-like enhancements.Conclusion.The proposed BPE classification tool for CEM achieves similar results than what is published in the literature for CE-MRI.

The features associated with mammography-occult MRI-detected newly diagnosed breast cancer analysed by comparing machine learning models with a logistic regression model

PURPOSE

To compare machine learning (ML) models with logistic regression model in order to identify the optimal factors associated with mammography-occult (i.e. false-negative mammographic findings) magnetic resonance imaging (MRI)-detected newly diagnosed breast cancer (BC).

MATERIAL AND METHODS

The present single-centre retrospective study included consecutive women with BC who underwent mammography and MRI (no more than 45 days apart) for breast cancer between January 2018 and May 2023. Various ML algorithms and binary logistic regression analysis were utilized to extract features linked to mammography-occult BC. These features were subsequently employed to create different models. The predictive value of these models was assessed using receiver operating characteristic curve analysis.

RESULTS

This study included 1957 malignant lesions from 1914 patients, with an average age of 51.64 ± 9.92 years and a range of 20-86 years. Among these lesions, there were 485 mammography-occult BCs. The optimal features of mammography-occult BC included calcification status, tumour size, mammographic density, age, lesion enhancement type on MRI, and histological type. Among the different ML models (ANN, L1-LR, RF, and SVM) and the LR-based combined model, the ANN model with RF features was found to be the optimal model. It demonstrated the best discriminative performance in predicting mammography false- negative findings, with an AUC of 0.912, an accuracy of 86.90%, a sensitivity of 85.85%, and a specificity of 84.18%.

CONCLUSION

Mammography-occult MRI-detected breast cancers have features that should be considered when performing breast MRI to improve the detection rate for breast cancer and aid in clinician management.

Generalisable deep learning method for mammographic density prediction across imaging techniques and self-reported race

BACKGROUND

Breast density is an important risk factor for breast cancer complemented by a higher risk of cancers being missed during screening of dense breasts due to reduced sensitivity of mammography. Automated, deep learning-based prediction of breast density could provide subject-specific risk assessment and flag difficult cases during screening. However, there is a lack of evidence for generalisability across imaging techniques and, importantly, across race.

METHODS

This study used a large, racially diverse dataset with 69,697 mammographic studies comprising 451,642 individual images from 23,057 female participants. A deep learning model was developed for four-class BI-RADS density prediction. A comprehensive performance evaluation assessed the generalisability across two imaging techniques, full-field digital mammography (FFDM) and two-dimensional synthetic (2DS) mammography. A detailed subgroup performance and bias analysis assessed the generalisability across participants' race.

RESULTS

Here we show that a model trained on FFDM-only achieves a 4-class BI-RADS classification accuracy of 80.5% (79.7-81.4) on FFDM and 79.4% (78.5-80.2) on unseen 2DS data. When trained on both FFDM and 2DS images, the performance increases to 82.3% (81.4-83.0) and 82.3% (81.3-83.1). Racial subgroup analysis shows unbiased performance across Black, White, and Asian participants, despite a separate analysis confirming that race can be predicted from the images with a high accuracy of 86.7% (86.0-87.4).

CONCLUSIONS

Deep learning-based breast density prediction generalises across imaging techniques and race. No substantial disparities are found for any subgroup, including races that were never seen during model development, suggesting that density predictions are unbiased.

Adoption of AI in Oncological Imaging: Ethical, Regulatory, and Medical-Legal Challenges

Artificial Intelligence (AI) algorithms have shown great promise in oncological imaging, outperforming or matching radiologists in retrospective studies, signifying their potential for advanced screening capabilities. These AI tools offer valuable support to radiologists, assisting them in critical tasks such as prioritizing reporting, early cancer detection, and precise measurements, thereby bolstering clinical decision-making. With the healthcare landscape witnessing a surge in imaging requests and a decline in available radiologists, the integration of AI has become increasingly appealing. By streamlining workflow efficiency and enhancing patient care, AI presents a transformative solution to the challenges faced by oncological imaging practices. Nevertheless, successful AI integration necessitates navigating various ethical, regulatory, and medical-legal challenges. This review endeavors to provide a comprehensive overview of these obstacles, aiming to foster a responsible and effective implementation of AI in oncological imaging.

A scoping review of programme specific mammographic breast density related guidelines and practices within breast screening programmes

INTRODUCTION

High mammographic breast density (MBD) is an independent breast cancer risk factor. In organised breast screening settings, discussions are ongoing regarding the optimal clinical role of MBD to help guide screening decisions. The aim of this scoping review was to provide an overview of current practices incorporating MBD within population-based breast screening programmes and from professional organisations internationally.

METHODS

This scoping review was conducted in accordance with the framework proposed by the Joanna Briggs Institute. The electronic databases, MEDLINE (PubMed), EMBASE, CINAHL Plus, Scopus, and Web of Science were systematically searched. Grey literature sources, websites of international breast screening programmes, and relevant government organisations were searched to identify further relevant literature. Data from identified materials were extracted and presented as a narrative summary.

RESULTS

The search identified 78 relevant documents. Documents were identified for breast screening programmes in 18 countries relating to screening intervals for women with dense breasts, MBD measurement, reporting, notification, and guiding supplemental screening. Documents were identified from 18 international professional organisations with the majority of material relating to supplemental screening guidance for women with dense breasts. Key factors collated during the data extraction process as relevant considerations for MBD practices included the evidence base needed to inform decision-making processes and resources (healthcare system costs, radiology equipment, and workforce planning).

CONCLUSIONS

This scoping review summarises current practices and guidelines incorporating MBD in international population-based breast screening settings and highlights the absence of consensus between organised breast screening programmes incorporating MBD in current breast screening protocols.

A Pictorial Exploration of Mammary Paget Disease: Insights and Perspectives

Mammary Paget disease (MPD) is a rare condition primarily affecting adult women, characterized by unilateral skin changes in the nipple-areolar complex (NAC) and frequently associated with underlying breast carcinoma. Histologically, MPD is identified by large intraepidermal epithelial cells (Paget cells) with distinct characteristics. Immunohistochemical profiles aid in distinguishing MPD from other skin conditions. Clinical evaluation and imaging techniques, including magnetic resonance imaging (MRI), are recommended if MPD is suspected, although definitive diagnosis always requires histological examination. This review delves into the historical context, epidemiology, pathogenesis, clinical manifestations, and diagnosis of MPD, emphasizing the need for early detection. The classification of MPD based on pathogenesis is explored, shedding light on its varied presentations. Treatment options, including mastectomy and breast-conserving surgery, are discussed with clear guidelines for different scenarios. Adjuvant therapies are considered, particularly in cases with underlying breast cancer. Prognostic factors are outlined, underlining the importance of early intervention. Looking to the future, emerging techniques, like liquid biopsy, new immunohistochemical and molecular markers, and artificial intelligence-based image analysis, hold the potential to transform MPD diagnosis and treatment. These innovations offer hope for early detection and improved patient care, though validation through large-scale clinical trials is needed.

Fibroglandular tissue segmentation in breast MRI using vision transformers: a multi-institutional evaluation

Accurate and automatic segmentation of fibroglandular tissue in breast MRI screening is essential for the quantification of breast density and background parenchymal enhancement. In this retrospective study, we developed and evaluated a transformer-based neural network for breast segmentation (TraBS) in multi-institutional MRI data, and compared its performance to the well established convolutional neural network nnUNet. TraBS and nnUNet were trained and tested on 200 internal and 40 external breast MRI examinations using manual segmentations generated by experienced human readers. Segmentation performance was assessed in terms of the Dice score and the average symmetric surface distance. The Dice score for nnUNet was lower than for TraBS on the internal testset (0.909 ± 0.069 versus 0.916 ± 0.067, P < 0.001) and on the external testset (0.824 ± 0.144 versus 0.864 ± 0.081, P = 0.004). Moreover, the average symmetric surface distance was higher (= worse) for nnUNet than for TraBS on the internal (0.657 ± 2.856 versus 0.548 ± 2.195, P = 0.001) and on the external testset (0.727 ± 0.620 versus 0.584 ± 0.413, P = 0.03). Our study demonstrates that transformer-based networks improve the quality of fibroglandular tissue segmentation in breast MRI compared to convolutional-based models like nnUNet. These findings might help to enhance the accuracy of breast density and parenchymal enhancement quantification in breast MRI screening.

Diagnostic accuracy of automated ACR BI-RADS breast density classification using deep convolutional neural networks

OBJECTIVES

High breast density is a well-known risk factor for breast cancer. This study aimed to develop and adapt two (MLO, CC) deep convolutional neural networks (DCNN) for automatic breast density classification on synthetic 2D tomosynthesis reconstructions.

METHODS

In total, 4605 synthetic 2D images (1665 patients, age: 57 ± 37 years) were labeled according to the ACR (American College of Radiology) density (A-D). Two DCNNs with 11 convolutional layers and 3 fully connected layers each, were trained with 70% of the data, whereas 20% was used for validation. The remaining 10% were used as a separate test dataset with 460 images (380 patients). All mammograms in the test dataset were read blinded by two radiologists (reader 1 with two and reader 2 with 11 years of dedicated mammographic experience in breast imaging), and the consensus was formed as the reference standard. The inter- and intra-reader reliabilities were assessed by calculating Cohen's kappa coefficients, and diagnostic accuracy measures of automated classification were evaluated.

RESULTS

The two models for MLO and CC projections had a mean sensitivity of 80.4% (95%-CI 72.2-86.9), a specificity of 89.3% (95%-CI 85.4-92.3), and an accuracy of 89.6% (95%-CI 88.1-90.9) in the differentiation between ACR A/B and ACR C/D. DCNN versus human and inter-reader agreement were both "substantial" (Cohen's kappa: 0.61 versus 0.63).

CONCLUSION

The DCNN allows accurate, standardized, and observer-independent classification of breast density based on the ACR BI-RADS system.

KEY POINTS

• A DCNN performs on par with human experts in breast density assessment for synthetic 2D tomosynthesis reconstructions. • The proposed technique may be useful for accurate, standardized, and observer-independent breast density evaluation of tomosynthesis.

Artificial Intelligence Applications in Breast Imaging: Current Status and Future Directions

Attempts to use computers to aid in the detection of breast malignancies date back more than 20 years. Despite significant interest and investment, this has historically led to minimal or no significant improvement in performance and outcomes with traditional computer-aided detection. However, recent advances in artificial intelligence and machine learning are now starting to deliver on the promise of improved performance. There are at present more than 20 FDA-approved AI applications for breast imaging, but adoption and utilization are widely variable and low overall. Breast imaging is unique and has aspects that create both opportunities and challenges for AI development and implementation. Breast cancer screening programs worldwide rely on screening mammography to reduce the morbidity and mortality of breast cancer, and many of the most exciting research projects and available AI applications focus on cancer detection for mammography. There are, however, multiple additional potential applications for AI in breast imaging, including decision support, risk assessment, breast density quantitation, workflow and triage, quality evaluation, response to neoadjuvant chemotherapy assessment, and image enhancement. In this review the current status, availability, and future directions of investigation of these applications are discussed, as well as the opportunities and barriers to more widespread utilization.

How do patients perceive the AI-radiologists interaction? Results of a survey on 2119 responders

PURPOSE

In this study we investigate how patients perceive the interaction between artificial intelligence (AI) and radiologists by designing a survey.

METHOD

We created a survey focused on the application of Artificial Intelligence in radiology which consisted of 20 questions distributed in three sections:Only completed questionnaires were considered for analysis.

RESULTS

2119 subjects completed the survey. Among them, 1216 respondents were over 60 years old, showing interest in AI even though they were not digital natives. Although >45% of the respondents reported a high level of education, only 3% said they were AI experts. 87% of respondents favored using AI to support diagnosis but would like to be informed. Only 10% would consult another specialist if their doctor used AI support. Most respondents (76%) said they would not feel comfortable if the diagnosis was made by the AI alone, highlighting the importance of the physician's role in the emotional management of the patient. Finally, 36% of respondents were willing to discuss the topic further in a focus group.

CONCLUSION

Patients' perception of the use of AI in radiology was positive, although still strictly linked to the supervision of the radiologist. Respondents showed interest and willingness to learn more about AI in the medical field, confirming how patients' confidence in AI technology and its acceptance is central to its widespread use in clinical practice.

Automatic and standardized quality assurance of digital mammography and tomosynthesis with deep convolutional neural networks

OBJECTIVES

The aim of this study was to develop and validate a commercially available AI platform for the automatic determination of image quality in mammography and tomosynthesis considering a standardized set of features.

MATERIALS AND METHODS

In this retrospective study, 11,733 mammograms and synthetic 2D reconstructions from tomosynthesis of 4200 patients from two institutions were analyzed by assessing the presence of seven features which impact image quality in regard to breast positioning. Deep learning was applied to train five dCNN models on features detecting the presence of anatomical landmarks and three dCNN models for localization features. The validity of models was assessed by the calculation of the mean squared error in a test dataset and was compared to the reading by experienced radiologists.

RESULTS

Accuracies of the dCNN models ranged between 93.0% for the nipple visualization and 98.5% for the depiction of the pectoralis muscle in the CC view. Calculations based on regression models allow for precise measurements of distances and angles of breast positioning on mammograms and synthetic 2D reconstructions from tomosynthesis. All models showed almost perfect agreement compared to human reading with Cohen's kappa scores above 0.9.

CONCLUSIONS

An AI-based quality assessment system using a dCNN allows for precise, consistent and observer-independent rating of digital mammography and synthetic 2D reconstructions from tomosynthesis. Automation and standardization of quality assessment enable real-time feedback to technicians and radiologists that shall reduce a number of inadequate examinations according to PGMI (Perfect, Good, Moderate, Inadequate) criteria, reduce a number of recalls and provide a dependable training platform for inexperienced technicians.

Mammography biomarkers of cardiovascular and musculoskeletal health: A review

Breast density (BD) and breast arterial calcifications (BAC) can expand the role of mammography. In premenopause, BD is related to body fat composition: breast adipose tissue and total volume are potential indicators of fat storage in visceral depots, associated with higher risk of cardiovascular disease (CVD). Women with fatty breast have an increased likelihood of hypercholesterolemia. Women without cardiometabolic diseases with higher BD have a lower risk of diabetes mellitus, hypertension, chest pain, and peripheral vascular disease, while those with lower BD are at increased risk of cardiometabolic diseases. BAC, the expression of Monckeberg sclerosis, are associated with CVD risk. Their prevalence, 13 % overall, rises after menopause and is reduced in women aged over 65 receiving hormonal replacement therapy. Due to their distinct pathogenesis, BAC are associated with hypertension but not with other cardiovascular risk factors. Women with BAC have an increased risk of acute myocardial infarction, ischemic stroke, and CVD death; furthermore, moderate to severe BAC load is associated with coronary artery disease. The clinical use of BAC assessment is limited by their time-consuming manual/visual quantification, an issue possibly solved by artificial intelligence-based approaches addressing BAC complex topology as well as their large spectrum of extent and x-ray attenuations. A link between BD, BAC, and osteoporosis has been reported, but data are still inconclusive. Systematic, standardised reporting of BD and BAC should be encouraged.