Jul 2026· Frontiers in Medical Technology· Vol 8· 0 citations· 43 references
Medicine
TL;DR
This study evaluates the performance of transfer learning-based MobileNetV2 and ResNet50 architectures for binary breast cancer classification using mammography images to achieve reliable and competitive performance for breast cancer detection under realistic evaluation conditions and improve the diagnostic accuracy of mammogram classification.
Abstract
Breast cancer is a leading cause of death for women worldwide. There is a critical need for an early and accurate diagnosis to improve survival rates. This study evaluates the performance of transfer learning-based MobileNetV2 and ResNet50 architectures for binary breast cancer classification using mammography images. To enhance channel-wise feature learning, the Squeeze-and-Excitation (SE) attention mechanism was integrated into the network and partial fine tuning was performed by only freezing the early layers of the model. Furthermore, gradient-weighted class activation mapping was applied to visualize the important regions of mammogram images responsible for the model’s predictions. The performance evaluation was carried out on the Contrast Limited Adaptive Histogram Equalization (CLAHE) binary, Digital Database for Screening Mammography (DDSM), INbreast, and Mammographic Image Analysis Society (MIAS) datasets using stratified k-fold cross-validation. The experimental results demonstrate that one of the proposed frameworks, MobileNetV2-SE, achieved mean classification accuracies of 99.14%, 93.53%, 92.35%, and 96.33% on the DDSM, INbreast, MIAS, and CLAHE-binary datasets, respectively. In comparison, the ResNet50-SE model attained mean accuracies of 100%, 88.89%, 71.60%, and 97.98% on the same datasets. The experimental results demonstrate that the frameworks achieve reliable and competitive performance for breast cancer detection under realistic evaluation conditions and improve the diagnostic accuracy of mammogram classification. This work contributes to the field of medical image analysis and computer-aided diagnosis. Future work should focus on applying advanced class balancing methods to improve diagnostic accuracy and model robustness in a variety of medical imaging applications.
Breast cancer remains a leading cause of cancer death that affects women worldwide, and the burden is felt most acutely in low resource settings where mammography access is scarce, and radiologist coverage is thin. Transfer learning-based deep learning models were evaluated for practical utility in breast cancer screening where imaging resources are restricted. The CBIS-DDSM dataset (Kaggle JPG version) was used for this study, with pathology labels mapped into a binary benign-versus-malignant classification scheme. The dataset consisted of 3,568 full mammograms and 3,461 ROI images from over 1,500 patients after quality control. 5-fold StratifiedGroupKFold cross-validation was used to prevent patient-level data leakage, which causes performance inflation in published studies, and ensured that no patient data combined training and validation datasets. The research evaluated two types of input data which included cropped region-of-interest lesion patches and complete mammogram images. InceptionV3 achieved the highest performance among ROI models by reaching ROC-AUC 0.821 and PR-AUC 0.775 and sensitivity 0.785. The study used full mammograms to evaluate ResNet50 performance which achieved ROC-AUC 0.838 and sensitivity 0.790 results. The study shows that transfer learning acts as a strong base for AI-assisted mammography triage systems in low-resource environments when patient level assessment continues throughout system development.
Zinat Abdulkadiri, Muhammad A. Suleiman, Joshua Abah· FUDMA Journal of Sciences· 0 citations
This work presents a leakage-controlled deep-learning framework for breast cancer classification using the CBIS-DDSM mammography archive. The proposed pipeline combines patient-level data partitioning before augmentation, a two-stage transfer-learning strategy based on ResNet50, and an Inter-View Attention Fusion (IVAF) module for adaptive fusion of paired craniocaudal (CC) and mediolateral oblique (MLO) feature maps. IVAF was modeled as a light-weighted convolutional gating strategy added after the last ResNet50 convolutional layer in order to create a weighted spatial-channel representation from the paired mammography images. In terms of the performance of the model under CBIS-DDSM held-out testing protocol, the entire model scored an accuracy of 97.12%, sensitivity of 96.44%, specificity of 97.68%, and AUC-ROC of 0.9876 based on the test results obtained on 6,117 images of 222 different patients. The average accuracy obtained using 100 random seeds was found to be 97.11% ± 0.18%.
Magy Makram, Alber S. Aziz, Mary Monir Saeid et al.· Scientific Reports· 0 citations
It is suggested that integrating advanced CNN architectures with radiomics features can significantly enhance the accuracy and reliability of both lesion segmentation and treatment prediction in breast cancer, potentially leading to better clinical outcomes.
Astha Karnwal, E. Dhamija, Pradeeba Sridar· Indian Journal of Radiology...· 0 citations
Breast cancer is a leading cause of cancer morbidity and mortality among women globally, emphasizing the need for accurate and timely diagnostic methods. A systematic but innovative two phases transfer learning based deep learning classification framework is developed using popular EfficientNetB7 architecture architecture for breast cancer classification. Breast ultrasound imaging proves a high degree of complexity to extract features with limited available medical datasets to address it and the proposed methodology attributes to it. We develop a framework which encompasses formalisation into components like data augmentation, progressive fine-tuning and adaptive learning rate optimization as methods for model generalisation. Experiments on the Breast Ultrasound Images (BUSI) dataset show that the model achieves best accuracy of over 91. 25% when classifying breast lesions into benign, malignant, and normal. It shows good discriminative abilities (validation accuracy: 93.62%) and a trained model converge well. We compare our method with the current ones and show significant advancements over all previous methods making our approach suitable for computer-aided diagnosis systems in clinical workflows.
Madishetti Kavya, G. Thirupati· International Journal of Sci...· 0 citations
Breast cancer is a significant cause of cancer-related deaths among women worldwide. Its early identification and screening are essential for improved patient outcomes and reduced mortality rates. Histopathological image analysis is considered as the gold standard for the diagnosis and prognosis of breast cancer. Nevertheless, the complexity of Whole-Slide Images (WSI) and their manual examination make this task time consuming, and prone to pathologist subjectivity. Recently, Deep Learning (DL) technology has achieved remarkable success in computer vision. However, their application still faces critical challenges in pathology analysis, including Region-of-Interest (RoI) scale variations, inter- and intra-class heterogeneity, diverse staining protocols, and the scarcity of annotated datasets. Furthermore, DL model’s findings are opaque and lack decision-level transparency. This study proposes a novel explainable multi-model DL framework for breast cancer classification leveraging histopathological images. The framework integrates Contrast Limited Adaptive Histogram Equalization (CLAHE) for image contrast enhancement, and diverse data augmentation to mitigate class imbalance and overfitting. Proposed architecture ensembles two branches, one employes DenseNet201 benefiting from Transfer Learning (TL) via ImageNet weights, while other utilizes a custom light weight attention based Hierarchal Feature Fusion (HFF) Network. DenseNet201 utilizes multilevel features to effectively tackle gradient vanishing issues and capture intricate feature representations, while HFF-Net, designed specifically for biomedical imaging, leverages HFF stem and multiscale feature extraction with Swish activation to enhance learning stability. Attention mechanism introduced within HFF-Net further refines the output features. Final feature vectors from the two branches are fused at the Global Average Pooling (GAP) layer, consolidating discriminative information. Experimental results on BRACS dataset demonstrate the proposed framework achieves 97.15% accuracy, 92.59% precision, 93.81% recall, and a 93.19% F1-score in screening tasks, while for grading tasks, it attains 84.08% accuracy, 83.39% precision, 83.64% recall, and 83.44% F1-score on 4391 test samples. Additionally, Gradient-Class Activation Mapping (Grad-CAM) saliency heatmap are generated for visual representation of proposed model’s choices, thereby increased transparency. The integration of these advanced techniques significantly enhances diagnostic reliability, addressing the challenges in histopathological image analysis.
Muhammad Nabeel Mehmood, Muhammad Hassaan Ashraf· Informatica· 0 citations
Breast cancer remains one of the most prevalent malignancies affecting women worldwide, and early, accurate detection through medical imaging is central to improving survival outcomes. This paper presents an end-to-end computer-aided diagnosis (CAD) framework that performs lesion segmentation, region-of-interest (ROI) extraction, tumor stage estimation, benign/malignant classification, and visual explainability across two complementary imaging modalities: digital mammography, represented by the Curated Breast Imaging Subset of the Digital Database for Screening Mammography (CBIS-DDSM), and breast ultrasound, represented by the Breast Ultrasound Images (BUSI) dataset. For each modality, a modality-specific U-Net architecture was trained to generate binary lesion masks, from which tumor coverage percentage and a rule-based disease stage were derived. Three classifier backbones were benchmarked per modality — ResNet50, VGG16, and a custom convolutional neural network (CNN) for CBIS-DDSM (benign vs. malignant), and ResNet50, EfficientNet-B0, and a custom CNN for BUSI (benign vs. malignant vs. normal) — with Gradient-weighted Class Activation Mapping (Grad-CAM) applied to the strongest classifiers for interpretability. The trained pipelines were deployed in a Django-based web application allowing end users to upload an image and receive segmentation, ROI, stage, classification, confidence, and Grad-CAM outputs in real time. Results show a pronounced modality gap: the U-Net segmentation network achieved a validation Dice coefficient of 0.462 on BUSI versus 0.247 on CBIS-DDSM, and the best classifier reached 84.07% test accuracy on BUSI (ResNet50) compared with 61.86% on CBIS-DDSM (ResNet50), with markedly lower malignant-class recall on mammography (37%) than on ultrasound (76%). These findings quantify how imaging modality, lesion conspicuity, and dataset scale materially affect both segmentation and diagnostic performance, and they underscore the necessity of transparent, modality-aware reporting — including recall/sensitivity for the malignant class, not just aggregate accuracy — when evaluating deep learning systems intended to support breast cancer screening.
Ishita Rana, D. Shah, D. Variya· Journal of Intelligent Decis...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
