Lung Nodule Detection and Classification using Machine Learning Techniques

As lung cancer is second most leading cause of death, early detection of lung cancer is became necessary in many computer aided dignosis (CAD) systems. Recently many CAD systems have been implemented to detect the lung nodules which uses Computer Tomography (CT) scan images [2]. In this paper, some image pre-processing methods such as thresholding, clearing borders, morphological operations (viz., erosion, closing, opening) are discussed to detect lung nodule regions ie, Region of Interest (ROI) in patient lung CT scan images. Also, machine learning techniques such as Support Vector Machine (SVM) and Convolutional Neural Network (CNN) has been discussed for classifying the lung nodules and non-nodules objects in patient lung ct scan images using the sets of lung nodule regions. In this study, Lung Image Database Consortium image collection (LIDC-IDRI) dataset having patient CT scan images has been used to detect and classify lung nodules. Lung nodule classification accuacy of SVM is 90% and that of CNN is 91.66%