Classification of brain hemorrhage computed tomography images using OzNet hybrid algorithm

摘要

Classification of brain hemorrhage computed tomography (CT) images providesa better diagnostic implementation for emergency patients. Attentively,each brain CT image must be examined by doctors. This situation is time-consuming,exhausting, and sometimes leads to making errors. Hence, we aim tofind the best algorithm owing to a requirement for automatic classification ofCT images to detect brain hemorrhage. In this study, we developed OzNethybrid algorithm, which is a novel convolution neural networks (CNN) algorithm.Although OzNet achieves high classification performance, we combineit with Neighborhood Component Analysis (NCA) and many classifiers: Artificialneural networks (ANN), Adaboost, Bagging, Decision Tree, K-NearestNeighbor (K-NN), Linear Discriminant Analysis (LDA), Na?ve Bayes and SupportVector Machines (SVM). In addition, Oznet is utilized for feature extraction,where 4096 features are extracted from the fully connected layer. Thesefeatures are reduced to have significant and informative features with minimumloss by NCA. Eventually, we use these classifiers to classify these significantfeatures. Finally, experimental results display that OzNet-NCA-ANNexcellent classifier model and achieves 100 accuracy with created Dataset2 from Brain Hemorrhage CT images.