Automatic Classification of Lung Tumors Using KNN, SVM and CNN

Abstract

A lung tumor classification system has been designed and developed. This work
presents a new approach to the automated classification of lung tumors based on texture
features, which separate lung tumor images from healthy tissues in computed tomography
images. The feature image used for the tumor classification consists of computed
tomography images. The application of the proposed method for tracking tumor is demonstrated to help pathologists distinguish its type of lung tumor. A classification with an
accuracy of 89%, 97% and 98%, has been obtained by k-nearest neighbor, support vector
machine and convolutional neural networks