Pulmonary Nodule Classification and Recognition Based on Sparse Representation Algorithm

Abstract

Classification and identification of pulmonary nodules is a key link in computer-aided diagnosis of lung tumors. However, the key problem in the classification and identification of pulmonary nodules is how to extract comprehensive and effective features. In response to this problem, this paper presents a classification and identification of pulmonary nodules based on sparse representation algorithm. The method is based on the lung nodule LIDC standard database to extract the texture features of nodules, Then, the multi-slice ROI feature of the same nodule is selected as the data set, but the data disaster is caused. However, the sparse representation can effectively reduce the large amount of redundant data and make the feature information more comprehensive and effective. Experimental results show that, while ensuring efficiency, the proposed method can effectively improve the classification accuracy of pulmonary nodules, and then assist doctors in clinical diagnosis.