Comparative Study of Transfer Learning with VGG16 and VGG19 for Breast Cancer Classification

Breast cancer is a significant health challenge, particularly in resource-constrained regions like Pakistan, where early detection remains a major hurdle. This study evaluates the effectiveness of transfer learning using VGG16 and VGG19 models for mammogram analysis, employing the CBIS-DDSM dataset. Both models underwent 5-fold cross-validation to ensure robust performance. The VGG16 model achieved an accuracy of 72.78%, while VGG19 demonstrated a slightly lower accuracy of 70.79% with data augmentation. The findings underscore the potential of transfer learning to enhance diagnostic accuracy, particularly in areas with limited healthcare resources. This research highlights the importance of adopting AI-based diagnostic tools to improve breast cancer detection and contribute to reducing mortality rates associated with late-stage diagnoses.