Prediction and Early Diagnosis of Breast Cancer Using Deep Learning

Abstract

Breast cancer is the most prevalent type of cancer and the leading cause of cancer-related fatalities globally. Early detection is crucial but often difficult, resulting in severe consequences for many individuals. This work employs deep learning, particularly the VGG-16 Convolutional Neural Network (CNN) framework, to identify and diagnose breast cancer in histopathology images. A comprehensive breast cancer image dataset was examined using Random Forest, ResNet-50, VGG-16, and a custom CNN model. The VGG-16 model attained an accuracy of 86.96% in distinguishing between benign and malignant tissues. The ResNet-50 model, when combined with a Random Forest classifier, achieved an accuracy of 99.93%, while another CNN architecture with a similar combination reached 95.95%. The standalone CNN model achieved 79.99% accuracy. The top model, ResNet-50 with Random Forest, also achieved a recall of 99% and a precision of 99.88%. These results suggest that ensemble learning techniques, such as combining CNNs with Random Forest classifiers, can sometimes surpass the performance of traditional deep learning models alone. The research also highlights the significance of tumor markers such as HER2, ER, PR, and Ki-67 in determining malignancy, helping to assess tumor aggressiveness and potential treatment responses. This study demonstrates that integrating deep learning and machine learning models can enhance the precision and clinical relevance of breast cancer diagnosis, thereby potentially reducing the global impact of the disease and saving lives.