Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{202779,
        author = {P. Jayaseelan and D. Anupriya},
        title = {Comparative Analysis of Densenet121 and Xception Networks for Breast Cancer Detection Using Mini-Mias Mammograms},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {9553-9569},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=202779},
        abstract = {Breast cancer is one of the leading causes of mortality among women worldwide, making early and accurate diagnosis essential for effective treatment and survival. Mammogram image analysis plays a significant role in Computer-Aided Diagnosis (CAD) systems for detecting breast abnormalities. However, manual interpretation of mammograms is challenging due to variations in breast tissue density, image noise, and low contrast regions. To address these challenges, this research proposes a comparative deep learning framework using DenseNet121 and Xception transfer learning models for breast cancer detection on the Mini-MIAS Database dataset. The proposed methodology includes preprocessing techniques such as ROI extraction, CLAHE, median filtering, histogram equalization, and image normalization to enhance mammogram quality. Data augmentation techniques including rotation, zooming, horizontal flipping, and brightness adjustment are applied to improve model generalization and reduce overfitting. Both models are trained using transfer learning with optimized hyperparameters for classification performance evaluation. The performance of DenseNet121 and Xception is analyzed using metrics such as accuracy, precision, recall, sensitivity, specificity, ROC curve, loss, and training time. The study aims to provide a low-computation and efficient breast cancer detection framework while comparing dense connectivity and depthwise separable convolution architectures. Experimental results are expected to demonstrate that DenseNet121 achieves slightly higher classification accuracy, whereas Xception provides faster computational performance for mammogram analysis.},
        keywords = {Breast Cancer Detection, Mammogram Classification, Deep Learning, Transfer Learning, DenseNet121, Xception, Mini-MIAS Dataset, Computer-Aided Diagnosis (CAD), Image Enhancement, Medical Image Processing},
        month = {May},
        }