STM Journals Peer-reviewed research from Consortium e-Learning Network

Research and Reviews: Journal of Oncology and Hematology

Breast Cancer Detection and Multiple Classification Using CNN

  1. Dhinakaran M.
  2. Ritik Lahoti
  3. Riyanshi Singh
  4. Saransh Chauhan
  5. Tanvi Sharma

Abstract

Although some efforts have been made in the form of preventative screening programs, breast cancer remains one of the rising causes of death in women. Computer-assisted diagnosis is needed because of the rapidly increasing number of mammograms that can be collected by these programs. Performance metrics are not significantly improved by computer aided detection methods designed to improve diagnosis without a large number of sequential readings. In this context, self-imaging techniques developed by deep learning show a possible direction to help diagnose breast cancer. In this research, we describe a deep learning method to classify breast cancer into several categories based on the Convolutional Neural Network (CNN) model. The proposed procedure attempts to classify breast tumors into benign or malignant categories; such as, fibroadenoma, lobular carcinoma, etc. In our experiments on histopathological images from the BreakHis database, the DenseNet CNN model outperformed the algorithm for multi-stage breast cancer, as anticipated, given the limited tumor sample. classification problem with an accuracy of 95.6%.

Keywords

Support