Enhancing Breast Cancer Detection of X-ray Imaging Using a Hybrid Canny Edge and Dense CNN Approach
Esraa Mohammed Alazzawi, Mohammed Sadoon Hathal · 2025
Breast cancer remains one of the leading causes of mortality among women worldwide, emphasizing the need for early and accurate detection. Traditional deep learning-based approaches for breast cancer classification often struggle with capturing fine structural details in X-ray images, leading to reduced diagnostic performance. This study proposes a hybrid approach that integrates Canny edge detection with a Dense Convolutional Neural Network (Dense CNN) to enhance breast cancer detection. The edge-aware filtering technique improves feature extraction by preserving critical structural information, thereby enhancing the model’s ability to distinguish malignant and benign regions. We evaluate the proposed Canny Edge + Dense CNN model against a baseline Dense CNN approach on a benchmark breast cancer dataset. Experimental results demonstrate that the Canny Edge + Dense CNN model achieves an accuracy of 88.37 %, significantly outperforming the Dense CNN, which attains 72.73 % accuracy. These findings highlight the effectiveness of incorporating edge-aware filtering techniques into deep learning architectures for improved breast cancer diagnosis, offering a promising direction for clinical applications.