Pre-Trained Deep Neural Network-Based Computer-Aided Breast Tumor Diagnosis Using ROI Structures

Deep neural network (DNN) based computer-aided breast tumor diagnosis (CABTD) method plays a vital role in the early detection and of tumors. However, Brightness mode (B-mode) ultrasound image derives training feature samples that make closer isolation toward infection part. Hence, it is expensive due to meta-heuristic search features occupying global region interest (ROI) structures input images. Thus, may lead high computational complexity pre-trained DNN-based CABTD method. This paper proposes novel ensemble using global- local-ROI-structures B-mode It conveys additional consideration for further enhancing method’s diagnostic performance without degrading its visual quality. The are extracted at various depths (18, 50, 101) from local ROI feed support vector machine better classification. From experimental results, has been observed combined structure small depth residual ResNet18 (0.8 %) produced significant improvement pixel ratio as compared ResNet50 (0.5 ResNet101 (0.3 %), respectively. Subsequently, methods have tested by influencing diagnose two specific tumors (Benign Malignant) improve accuracy (86%) Dense Net, Alex VGG Google Net. Moreover, reduces ResNet18,