This paper investigates computer aided diagnosis (CAD) techniques that allow detection of lung cancer through analysis of chest computed tomography (CT) images. In CAD two main problems have to be solved successfully in order to make the clinical acceptance of CAD systems a reality. One is the segmentation of the organ of interest, which in case of the lungs is already a challenge. Second is nodule detection, in which nodule features (like geometric properties, image intensity, shape and size) have to be taken into consideration. In the last few years, significant advancements have been made in this area. Researchers have presented many techniques far detection and classification of lung nodules, which show promising results. The paper aims at describing the overall structure for such a CAD system along with the review of different techniques, presented in recent years, for analyzing lung CT images for early detection of cancer.