An effective neural network model for lung nodule detection in CT images with optimal fuzzy model

Abstract

Cancer disease is assumed as a gathering of diseases which is initiated because of uncontrolled cell growth. An early analysis of Lung Nodules (LN) can possibly enhance the prognosis and in future can save numerous lives every year. In the proposed research work, the LN detection from the ELCAP lung image database is analyzed by image segmentation and classification techniques. Initially, the exact portion of the lung image is achieved and then it is subjected to pre-processing where the image contrast level is enhanced by the imadjust function of MATLAB. Next to that, the potential nodules are segmented by the Fuzzy C-Means (FCM) and then some features are extracted for effective classification. Based on the selected or extracted feature sets, the images are classified as two types (nodule detected and normal lung) by the proposed classifier i.e. Artificial Neural Network (ANN) with weight optimization. The performances of the proposed algorithm and classifier are tested on the chosen datasets in terms of sensitivity, specificity and accuracy. The results demonstrate that ANN with Oppositional based Ant Lion Optimization (OALO) algorithm achieves high accuracy and less execution time compared to existing algorithms.