Multiple Class Breast Cancer Detection Method Based on Deep Learning and MIRRCNN Model
G. Sajiv, G. Ramkumar
Abstract
The DCNN is a powerful deep learning system. Breast cancer classification, segmentation, and detection have previously been shown to be improved by DCNNs in a range of medical imaging modalities. As a global disease, breast cancer is one of the most frequent and most destructive. In this paper, the MIRRCNN model was employed to classify breast cancer based on recurrent residual convolutional neural networks (RNNs). Inception (Inception-v4), ResNet, and the Recurrent Convolutional Neural Network are only a few of the robust models in the DCNN family. Compared to Inception, Residual, and RCNN networks, the MIRRCNN is superior in object recognition. A publicly accessible database is utilised to classify breast cancer using the MIRRCNN method. Researchers compared their findings in terms of picture, patches, and patient-level categorization to previous machine learning and data mining methodologies. This model surpasses previous methods in terms of sensitivity, AUC, the ROC chart, and overall accuracy for both datasets.