Lung Cancer Classification using Curvelet Transform and Neural Network with Radial Basis Function

Abstract

In this paper, computed tomographic (CT) chest images were investigated to develop an automated system to discriminate lung cancer. These were done by analyzing Data recorded for patients with benign cancer, and also patients with malignant lung cancer were taken in account. The techniques for utilized feature extraction included features derived from texture analysis based on Gray Level Co-occurrence Matrix (GLCM) of the input image, as well as features derived from curvelet transform-based features. An artificial neural network (ANN) with radial basis function classifier was utilized to classify the type of cancer whether benign or malignant. The results have shown that using curvelet domain features gives the highest rate to recognize lung cancer. Classification correct rate is up to 96%. KeywordsCurvelet transform, lung cancer, image processing, ANN with radial basis function. Negative(Benign)1