Morphological Filter Aided GMM Technique for Lung Nodule Detection

Abstract

Lung cancer is one of the deadliest human disorders in all over the world. Early stage detection and identification of lung nodules from widely used High Resolution Computed Tomography (HRCT) images, helps in prevention of the disease. This paper is aimed to develop an automated computer-aided lung nodule detection system from HRCT images to provide a reliable second opinion to the radiologist and expert for further treatment. In this work a morphological filter aided Gaussian Mixture Model (GMM) is introduced for nodule segmentation and candidate detection. Support Vector Machine (SVM) with 10-fold cross validation technique is employed for nodule detection using LIDC/IDRI dataset. Finally, the reported work has detected the lung nodules with an overall sensitivity, specificity and accuracy of 89.77%, 86.92% and 88.24% respectively.