Extraction of lung lobes from CT scans by identifying incomplete fissure based on fissure, vessels and bronchi

In image processing the segmentation of image is a major task. The segmentation is related to clinical practice and challenging for severe diseases or incomplete fissures. At first, it detects the foreground and background of the lung image. After that the automated segmentation methods such as Marker based watershed transformation and Cellular automata technique are used to segment the lungs. In this transformation the lungs are divided into lobes. The lung image contains fissure, vessels and bronchi. From this segmentation result the distance between the incomplete fissures is calculated. Graph is drawn by computing the maximum and mean distance based on segmentation result. This segmentation can be analyzed by the integration of several anatomical structures against misfissures or incomplete fissures. For evaluation the method was compared to a recently published method on 20 CT scans with severe disease. Finally, analyze the relation between segment quality and incomplete fissure it shows the robust against incomplete fissures.