Investigation of Convolutional Neural Networks in the Tasks of Medical Images Analysis and Classification of Breast Tumors

Abstract

The problem of medical images analysis and classification of breast tumors is considered. For its solution the application of different convolutional neural networks CNN is suggested. The experimental investigations of the suggested CNN on the standard data set BreakHis were carried out in problems of binary and multi-class classification. The efficiency of different CNN was estimated and comparison with known results was performed. The best class of CNN for this problem was determined with the highest accuracy.