期刊名称：International Journal of Electronics Communication and Computer Technology
印刷版ISSN：2249-7838
出版年度：2012
卷号：2
期号：5
页码：208-212
出版社：International Journal of Electronics Communication and Computer Technology
摘要：Pulmonary Segmentation of lungs from chest CT images is useful in diagnosis of abnormalities and surgery planning. Lung is segmented from chest CT using iterative thresholding method. Further processing is carried out to visualize fissures characteristics and each seg ment inside the lung. Lung fissures are extracted using canny edge detection. Morphological operations are applied to smooth the lung borders. Lobe seg mentation is carried out by edge tracking and region filling algorithms to differentiate the segments. The method is evaluated on chest CT images of normal and abnormal cases. Confusion matrix is computed for determination of pulmona ry seg mentation accuracy. The segmentation method is automatic and shown good accuracy
