A Methodological Review on Breast Cancer Detection using Deep Learning in Histopathological images

Breast Cancer (BC) is one of the life threatening diseases which is strongly spread over human breast which causes death especially for women, and early precise diagnosis plays a significant role that can improve the patient’s outcome. It is significant to assist the early detection of breast cancer for medical involvement. Deep learning (DL) has turn out to be well- liked in medical imaging answers and has established various levels of performance in diagnosing cancerous images. The histopathological image analysis is a critical examination of BC and study the disease using tissue samples but suffers from subjectivity, complexity and limited Perfection. This methodological study analyzes in a comprehensive way and the state of the art(SOTA) of deep learning techniques for the diagnosis of BC. By using advanced deep learning models with accurate Convolution neural networks(CNN) ,Transfer learning and Vision Transfers (ViT), the process showcases how the automated and approximate detection of BC can be achieved with higher Perfection. Also this research focuses on detecting Breast cancer from Histopathological image dataset using deep learning models based on pre-trained ResNet50V2 approach. The proposed method achieved the Perfection of 75% while training and 74% on validation. This study projects to suggests the researchers and clinicians in understanding the nature of BC, improving diagnostic Perfection, and path the way toward effective treatment methods. Ultimately, it fosters a unified understanding of DL advancements in histopathological imaging and proposes innovation in early BC detection..