Breast Cancer Classification On Mammogram Images

Breast cancer is a leading cause of illness and death worldwide, with early detection crucial for improving survival rates. Recent advancements in deep learning have revolutionized breast cancer detection, enabling highly precise and effective diagnostic techniques. This literature analysis consolidates information from 20 pivotal experiments. emphasizing datasets, deep learning architectures, preprocessing techniques, assessment measures, and resultant outcomes. Widely used datasets such as CBIS-DDSM, INbreast, MIAS, and BreaKHis form the basis for many studies, while custom and Private datasets address specific challenges. Architectures including EfficientNet, DenseNet, ResNet, Vision Transformers, and ensemble methods demonstrate superior performance, with reported accuracies often exceeding 95 %. Methods such as data augmentation, picture segmentation, and explainable AI improves model resilience and clarity of interpretation. Nonetheless, challenges including dataset reliance, computational complexity, and generalization difficulties persist. significantly. This analysis delineates the strengths and weaknesses of existing methodologies, providing insights into prospective avenues, such as the necessity for diversified datasets, resourceefficient models, and clinically verified solutions. By tackling these issues, deep learning might substantially enhance breast cancer diagnosis and elevate patient outcomes.