Computer Aided Fracture Diagnosis Based on Integrated Learning

Abstract

Fracture refers to the complete or partial rupture of bone structure, which requires accurate diagnosis by orthopedic surgeons and treatment methods. Therefore, it is of great significance to study automatic fracture detection. In this paper, in order to solve the problem of low accuracy of fracture image determination caused by incomplete feature extraction of traditional features, a deep learning method is used to build a convolutional neural network model framework, and a fracture image detection method based on deep features and integrated learning is proposed. Using data enhancement to preprocess the MURA image data set, when extracting image features, Alexnet is used as a feature extractor to obtain sufficiently effective image features, and when training the classifier, the idea of integrated learning in machine learning is adopted. Train the classifiers after feature extraction, and give them different weight values according to the contribution of each classifier, to achieve better performance than a single classifier, and improve the accuracy of image classification. The experimental results of this method on the MURA data set show its good classification performance.