Search Results (90)

Introduction: An inverse association between high mammographic breast density (MBD) and pathologic complete response (pCR) following neoadjuvant chemotherapy (NAC) for early breast cancer (EBC) has been reported. However, the relationship of MBD to relapse-free (RFS) and breast cancer-specific survival (BCSS) is unexplored. This study aims to validate the relationship between MBD and NAC pCR in EBC and to assess correlations with RFS and BCSS. Materials & Methods: MBD was measured on contralateral mammograms in 127 women before NAC using Cumulus software. The percent dense area was correlated with patient and tumour characteristics, pCR, RFS and BCSS. Results: Mean MBD was higher in relapsing patients (p = 0.041) but did not vary by pCR or BC-deaths. As a dichotomous variable, no difference was seen between high and low MBD cohorts for pCR (17.5 vs. 25.0%, p = 0.15), BC relapse (38 vs. 30%, p = 0.15) or BC death (32 vs. 25%, p = 0.20). A planned analysis by body mass index (BMI) demonstrated high MBD associated with lower pCR (0% vs. 28.1%, p = 0.036) and trends for higher relapse (56% vs. 28%, p = 0.063) and BC deaths (56 vs. 28%, (p = 0.071)) in obese patients. No relationship was observed in non-obese patients. Conclusions: Obesity and high MBD may interact to cause chemoresistance. Further research in these patients is warranted. Full article

Background/Objectives: Artificial intelligence (AI)-based systems are increasingly being used to assist radiologists in detecting breast cancer on mammograms. However, applying fixed AI score thresholds across diverse lesion types may compromise diagnostic performance, especially in women with dense breasts. This study aimed to determine optimal category-specific AI thresholds and to analyze discrepancies between AI predictions and radiologist assessments, particularly for BI-RADS 4A versus 4B/4C lesions. Methods: We retrospectively analyzed 194 mammograms (76 BI-RADS 4A and 118 BI-RADS 4B/4C) using FDA-approved AI software. Lesion characteristics, breast density, AI scores, and pathology results were collected. A receiver operating characteristic (ROC) analysis was conducted to determine the optimal thresholds via Youden’s index. Discrepancy analysis focused on BI-RADS 4A lesions with AI scores of ≥35 and BI-RADS 4B/4C lesions with AI scores of <35. Results: AI scores were significantly higher in malignant versus benign cases (72.1 vs. 20.9; p < 0.001). The optimal AI threshold was 19 for BI-RADS 4A (AUC = 0.685) and 63 for BI-RADS 4B/4C (AUC = 0.908). In discordant cases, BI-RADS 4A lesions with scores of ≥35 had a malignancy rate of 43.8%, while BI-RADS 4B/4C lesions with scores of <35 had a malignancy rate of 19.5%. Conclusions: Using category-specific AI thresholds improves diagnostic accuracy and supports radiologist decision-making. However, limitations persist in BI-RADS 4A cases with overlapping scores, reinforcing the need for radiologist oversight and tailored AI integration strategies in clinical practice. Full article

Background/Objectives: Colorectal (CRC), breast (BC), and cervical cancer (CC) pose a significant health burden, yet screening programs have been proven to reduce cancer-specific mortality and other non-lethal endpoints. Mobile health (mHealth) technologies can enhance adherence, but effectiveness varies. This scoping review aims to explore mHealth apps for cancer screening developed with community engagement, identifying research approaches and gaps. Methods: A scoping review following PRISMA-ScR guidelines analyzed studies on mHealth apps for cancer screening developed through community engagement. Community engagement was classified per WHO’s definition. Databases were searched using a PCC-based strategy; eligible studies involved app development, excluding hypothetical apps or text messaging-/social media-only interventions. Screening and data extraction were conducted independently. Results: Thirteen articles were included. Findings indicate a growing but limited body of evidence, with most studies focusing on CRC and BC and involving minority populations through mHealth apps. Key engagement phases included research design, CAB establishment, and recruitment, while priority setting was never community-led. The wMammogram, Meet ALEX, and mMammogram apps improved screening knowledge, intention, and participation, while ColorApp enhanced knowledge but not attitudes. Only CBPR-based studies included dissemination, and one involved the CAB in data analysis. Some studies acknowledged community contributions, though details on ColorApp’s engagement were limited. Conclusions: Standardized engagement frameworks combined with mHealth were associated with greater community involvement and may improve equity. No community-designed mHealth app was found for CC screening, despite its relevance. Future research should address gaps in CC programs, prioritize early community involvement, and assess the long-term impact of mHealth interventions. Full article

Background/Objectives: Digital mammography is widely used for breast cancer screening; however, variations in system design and automatic exposure control (AEC) strategies can lead to significant differences in radiation dose, potentially affecting the diagnostic quality and patient safety. In this study, we aimed to determine the effect of various mammographic technologies on the in vivo mean glandular doses (MGDs) that are received in clinical settings. Methods: The MGDs and applied acquisition parameters from 194,608 mammograms, acquired employing AEC using different digital mammography systems (GE, Siemens, and two different models of Hologic), were retrospectively collected. The potential variation in MGD resulting from different technologies (system and target/filter combination) was assessed employing the Kruskal–Wallis test, followed by Dunn’s post hoc. The AEC optimization of acquisition parameters (kVp, mAs) within each system was investigated through a multi-regression analysis as a function of the compressed breast thickness (CBT). The trend line of these parameters in addition to the MGD and source-to-breast distance were also plotted and compared. Results: There were significant variations in delivered doses per CBT based on which technology was used (p < 0.001). The regression analyses revealed system-specific differences in AEC adjustments of mAs and kVp in response to CBT changes. As the CBT increases, the MGD increases with different degrees, rates, and patterns across systems due to differences in AEC strategies. Conclusions: The MGD is affected by the applied technology, which is different between systems. Clinicians need to be aware of these variations and how they affect the MGD. Additionally, manufacturers may need to consider standardizing the implemented technology effects on the MGDs. Full article

Background: Breast cancer is the second leading cause of cancer-related mortality among women, accounting for 12% of cases. Early diagnosis, based on the identification of radiological features, such as masses and microcalcifications in mammograms, is crucial for reducing mortality rates. However, manual interpretation by radiologists is complex and subject to variability, emphasizing the need for automated diagnostic tools to enhance accuracy and efficiency. This study compares a radiomics workflow based on machine learning (ML) with a deep learning (DL) approach for classifying breast lesions as benign or malignant. Methods: matRadiomics was used to extract radiomics features from mammographic images of 1219 patients from the CBIS-DDSM public database, including 581 cases of microcalcifications and 638 of masses. Among the ML models, a linear discriminant analysis (LDA) demonstrated the best performance for both lesion types. External validation was conducted on a private dataset of 222 images to evaluate generalizability to an independent cohort. Additionally, a deep learning approach based on the EfficientNetB6 model was employed for comparison. Results: The LDA model achieved a mean validation AUC of 68.28% for microcalcifications and 61.53% for masses. In the external validation, AUC values of 66.9% and 61.5% were obtained, respectively. In contrast, the EfficientNetB6 model demonstrated superior performance, achieving an AUC of 81.52% for microcalcifications and 76.24% for masses, highlighting the potential of DL for improved diagnostic accuracy. Conclusions: This study underscores the limitations of ML-based radiomics in breast cancer diagnosis. Deep learning proves to be a more effective approach, offering enhanced accuracy and supporting clinicians in improving patient management. Full article

Breast cancer is the most common disease in women, with 287,800 new cases and 43,200 deaths in 2022 across United States. Early mammographic picture analysis and processing reduce mortality and enable efficient treatment. Several deep-learning-based mammography classification methods have been developed. Due to low-contrast images and irrelevant information in publicly available breast cancer datasets, existing models generally perform poorly. Pre-trained convolutional neural network models trained on generic datasets tend to extract irrelevant features when applied to domain-specific classification tasks, highlighting the need for a feature selection mechanism to transform high-dimensional data into a more discriminative feature space. This work introduces an innovative and effective multi-step pathway to overcome these restrictions. In preprocessing, mammographic pictures are haze-reduced using adaptive transformation, normalized using a cropping algorithm, and balanced using rotation, flipping, and noise addition. A 32-layer convolutional neural model inspired by YOLO, U-Net, and ResNet is intended to extract highly discriminative features for breast cancer classification. A modified Grey Wolf Optimization algorithm with three significant adjustments improves feature selection and redundancy removal over the previous approach. The robustness and efficacy of the proposed model in the classification of breast cancer were validated by its consistently high performance across multiple benchmark mammogram datasets. The model’s constant and better performance proves its robust generalization, giving it a powerful solution for binary and multiclass breast cancer classification. Full article

The early and accurate detection of breast lesions through mammography is crucial for improving survival rates. However, the existing deep learning-based methods often rely on costly pixel-level annotations, limiting their scalability in real-world applications. To address this issue, a novel local extremum mapping (LEM) mechanism is proposed for mammogram classification and weakly supervised lesion localization. The proposed method first divides the input mammogram into multiple regions and generates score maps through convolutional neural networks. Then, it identifies the most informative regions by filtering local extrema in the score maps and aggregating their scores for final classification. This strategy enables lesion localization with only image-level labels, significantly reducing annotation costs. Experiments on two public mammography datasets, CBIS-DDSM and INbreast, demonstrate that the proposed method achieves competitive performance. On the INbreast dataset, LEM improves classification accuracy to 96.3% with an AUC of 0.976. Furthermore, the proposed method effectively localizes lesions with a dice similarity coefficient of 0.37, outperforming Grad-CAM and other baseline approaches. These results highlight the practical significance and potential clinical applications of our approach, making automated mammogram analysis more accessible and efficient. Full article

Background/Objectives: The primary purpose of this study is to provide a more in-depth insight into various demographic, clinical, and lifestyle factors in relation to breast cancer and to predict the extent to which certain variables described as “predictors” might lead to further investigation. By analyzing a large cohort, we are able to provide valuable and up-to-date information on breast cancer screening, support breast specialists, and further enhance international screening guidelines. Methods: We screened for breast cancer in a population of women aged 50 to 69 years by using the standardized breast cancer imaging screening method (breast mammography) and ultrasonography as a complementary imagistic tool, and we compared the results with the gold standard, breast biopsy. For this, 58,760 women with no known history of breast cancer coming from 4 major regions of Romania (North-East, North-West, South-East, and West) were first evaluated through mammography. Out of these, 3197 women with positive mammograms subsequently underwent a breast ultrasound examination. The remaining 688 patients with positive breast ultrasound were further referred for a breast biopsy. Results: The statistical analysis revealed several predictors such as the body mass index (BMI), positive family medical history of breast cancer, age at first birth, and age at menopause that influenced the progression from mammography (first stage of the screening program) towards echography (additional imaging modality). Furthermore, we established that age, age at first birth, and BMI are significant predictors of progression from echography towards biopsy (the last stage of the screening program). Furthermore, by analyzing the number of positive biopsies (688) out of the total number of patients in the study (58,760), we calculated a total breast cancer detection rate of 8 per 1000 patients. Lastly, by studying the patient demographics in the context of breast cancer (BC) screening, we observed that participants coming from an urban environment presented a higher rate of positive mammographic results as compared to ones of rural provenience. Conclusions: Our study analyzed a large cohort of patients and offers real world data which shows that multiple factors were positively associated with an increased risk of BC. Older age, older age at first birth, and an older menopausal age are all estrogen-dependent risk factors that were linked with an increased breast cancer risk in our study. Furthermore, our findings concerning the rural/urban disparities and regional differences highlight the need for region-specific interventions to address lifestyle risk factors, improve healthcare access, and enhance breast cancer screening and follow-up protocols, particularly in underserved areas like the North-East and South-East regions. Full article

Background: Breast cancer screening through mammography interpretation is crucial for early detection and improved patient outcomes. However, the manual classification of mammograms using the BIRADS (Breast Imaging-Reporting and Data System) remains challenging due to subtle imaging features, inter-reader variability, and increasing radiologist workload. Traditional computer-aided detection systems often struggle with complex feature extraction and contextual understanding of mammographic abnormalities. To address these limitations, this study proposes MammoViT, a novel hybrid deep learning framework that leverages both ResNet50’s hierarchical feature extraction capabilities and Vision Transformer’s ability to capture long-range dependencies in images. Methods: We implemented a multi-stage approach utilizing a pre-trained ResNet50 model for initial feature extraction from mammogram images. To address the significant class imbalance in our four-class BIRADS dataset, we applied SMOTE (Synthetic Minority Over-sampling Technique) to generate synthetic samples for minority classes. The extracted feature arrays were transformed into non-overlapping patches with positional encodings for Vision Transformer processing. The Vision Transformer employs multi-head self-attention mechanisms to capture both local and global relationships between image patches, with each attention head learning different aspects of spatial dependencies. The model was optimized using Keras Tuner and trained using 5-fold cross-validation with early stopping to prevent overfitting. Results: MammoViT achieved 97.4% accuracy in classifying mammogram images across different BIRADS categories. The model’s effectiveness was validated through comprehensive evaluation metrics, including a classification report, confusion matrix, probability distribution, and comparison with existing studies. Conclusions: MammoViT effectively combines ResNet50 and Vision Transformer architectures while addressing the challenge of imbalanced medical imaging datasets. The high accuracy and robust performance demonstrate its potential as a reliable tool for supporting clinical decision-making in breast cancer screening. Full article

Background: This study compares system-reported organ doses (ODs) to manually calculated mean glandular doses (MGDs) in mammography across multiple centers and manufacturers in Dubai. Methods: A retrospective study of 2754 anonymized mammograms from six clinics in Dubai were randomly retrieved from a central dose survey database. Organ doses were documented along with other dosimetry information like kVp, mAs, filter, target, compression force, and breast thickness. Mean glandular doses, MGDs, were calculated manually for all the patients using the Dance formula and inferential statistical analyses were run to compare the two figures and verify the factors affecting each. Results: Our study’s analysis revealed that manually calculated mean glandular doses (MGDs) provide a more reliable indicator of radiation exposure than organ doses (ODs) reported by DICOM, particularly in multi-vendor scenarios. Manually calculated MGD values were consistently lower than system-reported ODs (MLO view: 0.96 ± 0.37 mGy vs. 1.38 ± 0.45 mGy; CC view: 0.81 ± 0.33 mGy vs. 1.22 ± 0.38 mGy). Significant differences in both system-reported ODs and manually calculated MGDs were observed across centers (p < 0.001). Strong correlations between system-reported ODs and manually calculated MGDs were found for Siemens equipment (r = 0.923, p < 0.001) but only moderate correlations for GE systems (r = 0.638, p < 0.001). Calculated MGD values were significantly higher for GE equipment compared to Siemens (1.49 ± 0.77 mGy vs. 0.93 ± 0.33 mGy, p < 0.001). Conclusions: This study addresses concerns regarding mammography dosimetry accuracy by demonstrating the superiority of mean glandular doses over DICOM-generated organ doses. These findings empower practitioners to optimize dose levels, ensuring safer and more effective breast cancer screening protocols. Full article

Objective: Metabolic dysfunction-associated steatotic liver disease (MASLD), previously known as the most common chronic liver disease, is soon to be the leading indication for liver transplantation; however, the diagnosis may remain occult for decades. There is a need for biomarkers that identify patients at risk for MASLD and patients at risk for disease progression to optimize patient management and outcomes. Lymph node adiposity (LNA) is a novel marker of adiposity identified within axillary lymph nodes on screening mammography. Recent studies have demonstrated a correlation between LNA and cardiometabolic disease and cardiovascular disease risk. This study aimed to investigate the association between MASLD and LNA to evaluate the potential of mammographic LNA to serve as an imaging biomarker of MASLD. Methods: We identified women with pathology-proven MASLD who had a liver biopsy and a screening mammogram within 12 months of the liver biopsy. This resulted in a sample size of 161 women for final analysis that met the inclusion criteria. We evaluated lymph node adiposity through multiple measurements of the largest axillary lymph node visualized on mammography and correlated LNA with MASLD histology. Statistical analysis using univariable and multivariable logistic regression and odds ratios was performed using R version 4.1.0 (2021), the R Foundation for Statistical Computing Platform. Results: We found a significant association between MASLD and mammographic LNA, defined as lymph node (LN) length > 16 mm (p = 0.0004) that remained significant after adjusting for clinical factors, including body mass index (BMI). We additionally found a significant association between LNA and metabolic dysfunction-associated steatohepatitis (MASH), identified via liver biopsy (p = 0.0048). Conclusions: Mammographic lymph node adiposity may serve as a helpful imaging biomarker of MASLD in women who have an elevated risk for the development of MASH. Full article

Identifying breast masses is relevant in early cancer detection. Automatic identification using computational methods helps assist medical experts with this task. Although high values have been reported in breast mass classification from digital mammograms, most results have focused on a general benign/malignant classification. According to the BI-RADS standard, masses are associated with cancer risk by grade depending on their specific shape, margin, and density characteristics. This work presents a methodology of testing several descriptors on the INbreast dataset, finding those better related to clinical assessment. The analysis provides a description based on BI-RADS for mass classification by combining neural networks and image processing. The results show that masses associated with grades BI-RADS-2 to BI-RADS-5 can be identified, reaching a general accuracy and sensitivity of $0.88\pm 0.07$. While this initial study is limited to a single dataset, it demonstrates the possibility of generating a description for automatic classification that is directly linked to the information analyzed by medical experts in clinical practice. Full article

Breast cancer remains one of the primary causes of cancer-related deaths among women globally. Early detection via mammography is essential for improving prognosis and survival rates. However, mammogram diagnostic accuracy is severely hindered by dense breast tissue, which can obstruct potential malignancies, complicating early detection. To tackle this pressing issue, this study introduces an innovative approach that leverages Generative Adversarial Networks (GANs), specifically CycleGAN and GANHopper, to transform breast density in mammograms. The aim is to diminish the masking effect of dense tissue, thus enhancing the visibility of underlying malignancies. The method uses unsupervised image-to-image translation to gradually alter breast density (from high (ACR-D) to low (ACR-A)) in mammographic images, detecting obscured lesions while preserving original diagnostic features. We applied this approach to multiple mammographic datasets, demonstrating its effectiveness in diverse contexts. Experimental results exhibit substantial improvements in detecting potential malignancies concealed by dense breast tissue. The method significantly improved precision, recall, and F1-score metrics across all datasets, revealing previously obscured malignancies and image quality assessments confirmed the diagnostic relevance of transformed images. The study introduces a novel mammogram analysis method using advanced machine-learning techniques, enhancing diagnostic accuracy in dense breasts and potentially improving early breast cancer detection and patient outcomes. Full article

This study evaluates the potential of an AI-based computer-aided detection (AI-CAD) system in digital mammography for predicting pathologic complete response (pCR) in breast cancer patients after neoadjuvant chemotherapy (NAC). A retrospective analysis of 132 patients who underwent NAC and surgery between January 2020 and December 2022 was performed. Pre- and post-NAC mammograms were analyzed using conventional CAD and AI-CAD systems, with negative exams defined by the absence of marked abnormalities. Two radiologists reviewed mammography, ultrasound, MRI, and diffusion-weighted imaging (DWI). Concordance rates between CAD and AI-CAD were calculated, and the diagnostic performance, including the area under the receiver operating characteristics curve (AUC), was assessed. The pre-NAC concordance rates were 90.9% for CAD and 97% for AI-CAD, while post-NAC rates were 88.6% for CAD and 89.4% for AI-CAD. The MRI had the highest diagnostic performance for pCR prediction, with AI-CAD performing comparably to other modalities. Univariate analysis identified significant predictors of pCR, including AI-CAD, mammography, ultrasound, MRI, histologic grade, ER, PR, HER2, and Ki-67. In multivariable analysis, negative MRI, histologic grade 3, and HER2 positivity remained significant predictors. In conclusion, this study demonstrates that AI-CAD in digital mammography shows the potential to examine the pCR of breast cancer patients following NAC. Full article

Background: Breast density impacts upon breast cancer risk and recurrence, but its influence on breast cancer-specific survival is unclear. This study examines the influence of mammographic breast density (MBD) at diagnosis on breast cancer-specific survival. Methods: The data of 224 patients diagnosed with breast cancer were analyzed. Two area-based MBD measurement tools—AutoDensity and LIBRA—were used to measure MBD via a mammogram of the contralateral breast acquired at the time of diagnosis. These patients were split into two groups based on their percent breast density (PBD): high (PBD ≥ 20%) versus low (PBD < 20%). Breast cancer-specific survival in each of these PBD groups was assessed at a median follow-up of 34 months using Kaplan–Meier analysis and the Cox proportional hazards model. Results: The proportion of women with low PBD who died from breast cancer was significantly higher than that seen with high PBD (p = 0.01). The 5-year breast cancer-specific survival was poorer among women with low PBD than those with high PBD (0.348; 95% CI: 0.13–0.94) vs. 0.87; 95% CI: (0.8–0.96); p < 0.001)]. Women with higher breast density demonstrated longer survival regardless of the method of PBD measurement: LIBRA [log-rank test (Mantel–Cox): 9.4; p = 0.002)]; AutoDensity [log-rank test (Mantel–Cox) 7.6; p = 0.006]. Multivariate analysis also demonstrated that there was a higher risk of breast cancer-related deaths in women with low PBD (adjusted HR: 5.167; 95% CI: 1.974–13.521; p = 0.001). Conclusion: Women with <20% breast density at breast cancer diagnosis demonstrate poor survival regarding the disease. The impact of breast density on survival is not influenced by the method of measurement. Full article