CT Images Classification of Hepatic Hydatid Disease by Different Methods

Abstract

This paper discussed the application of traditional machine learning and deep learning methods in CT image classification of liver hydatid. At first, the Median filtering algorithm was used to preprocess the CT images, and three kinds of features were extracted. The decision tree(C4.5) and the support vector machine(SVM) Classifier were applied to classify and were compared with each other. we got that the SVM has higher classification performance for CT images of hepatic hydatid disease than decision tree(C4.5). Secondly, we set up the improved convolutional neural network(CNN) model which was used to classify the images and the classification accuracy was compared with the traditional machine learning. The accuracy of CNN is higher than traditional machine learning classifiers which were applied above. Our experimental results manifest that the application of deep learning lays a foundation for the later development of computer aided diagnosis system for CT-image of hepatic hydatid disease which offers more accurate diagnosis to assist doctors.