Performance evaluation of CAD system for Lung Cancer Detection

Lung Cancer is the foremost cause of eternal rest all around the world. Lung nodules are possible symptom of lung cancer and the early revealing helps in early treatment and enhance patient’s chances for endurance. An automated lung nodule segmentation of a region of interest (ROI) is the most challenging problem in clinical practices. For this basis, CAD systems for lung cancer detection have been intended in many dissertations. This paper reviews some of the current segmentation algorithms and techniques and also afford a comparative analysis of the accomplishment of the existing approaches. The LIDC dataset is used for performance analysis of segmentation algorithms. The experimental analysis shows that watershed algorithm has high accuracy when compared with other algorithms.