A Deep Learning Approach for Automated Breast Cancer Detection

Key Points

• To develop a deep learning-based system for detecting breast cancer using mammogram images.
• Used CBIS-DDSM mammogram images for analysis.
• Applied preprocessing techniques such as ROI extraction.
• Implemented transfer learning with models including ResNet50, MobileNet, and VGG16.
• Achieved high classification accuracy of approximately 97%.
• Demonstrated effectiveness for computer-aided diagnosis.