Early detection of breast cancer using robust back propagation neural network classifier

Abstract

Breast cancer is the most widely recognized malignancy in women and is the second most common and important cause of cancer death in women. Currently, there is no e􀵵 ective approach to avoid breast cancer, as its motivation is not yet fully understood. The common problems of women in recent years are breast cancer. The problem of breast cancer classification has been the main problem in recent years. This article gives a reasonable idea of the readiness of the mammogram image to find out the area a􀵵 ected by cancer, which is a vital advance in breast cancer detection. Therefore, the research focuses on examining and analyzing the image processing of cancerous diseases, especially breast cancer. When detecting breast cancer, images are obtained from a mammogram and subjected to treatment. The artificial neural network (ANN) plays an important role in this regard, in particular in the application of biomedical image processing for the early diagnosis of breast cancer. It is actually detected by these steps. Preprocessing, image splitting, attribute extraction and categorization. An average vector filter is used to perform the preprocessing operation. Helps eliminate noise and turn RGB image into grayscale image. Vessel segmentation is used to segment the image. To perform entity extraction, the technique of the Binary Local Scale Model (MLBP) method is used and optimized using the Robust Propagation Neural Network Classifier (RBPNN). Using MLBP, 12 functional derivatives were extracted. The extraction is performed on the basis of color, texture and shape. It plays an important role in detection. Finally, for the classi􀂿 cation of the extracted function, RBPNN is performed and then analyzes the test data with the trained data. To prove its usefulness, it produces great precision compared to other current games. The accuracy of this suggested analysis is 97%.