Future fusion+: Breast cancer tissue identification and early detection of deep hybrid featured based healthcare system

Abstract

The exponential increase in cancer cases, particularly breast cancer (BC), has prompted academics and business to develop more effective and trustworthy methods for classifying and identifying BC tissues. In contrast to traditional machine learning (ML) techniques, this work presents the development of a fusion of features AlexResNet+: a deep hybrid feature-based system of early BC detection in healthcare tissue identification. For deep feature extraction, we employed three of the most popular and effective deep learning models, AlexNet, ResNet50 and ResNet101. We employed ResNet50 with modified layered architectures while using AlexNet with five convolutional layers in order to maintain high dimensional deep features while maintaining the best computational efficiency. As a result of combining the deep features from the AlexNet and ResNet DL models, we were able to perform two-class classification using the Support Vector Machine with Radial Basis Function (SVM-RBF). Performances for AlexNet, shorted ResNet50, and hybrid features were collected separately to evaluate the effectiveness of the various feature sets. The counseled hybrid deep capabilities (AlexResNet+)-primarily based version has a most class accuracy of 95.87%, precision of 0.9760, sensitivity of 1.0, specificity of 0.9621, F-Measure of 0.9878 and AUC of 0.960, according to simulation results using DDMS mammography breast cancer tissue pictures.