![]() | Only 14 pages are availabe for public view |
Abstract Lung cancer is one of the most serious cancers in the world with the minimum survival rate. Lung nodules may be isolated from (solitary) or attached to (juxtapleural) other. In this paper a Computer Aided Diagnosis system is proposed to classify between solitary nodule and juxtapleural nodule inside the lungs.Two main auto-diagnostic schemes of supervised learning for classification are achieved.Three segmentation approaches are proposed.The three classifiers of the first scheme are K-Nearest Neighborhood, Artificial Neural Network and Support Vector Machine. In the second scheme, Deep Convolutional neural networksare used. Because of limited data sample and imbalanced data, 10-fold cross validation and random oversampling are used.The 3D reconstruction of pulmonary nodules based on the surface rendering technique and visualization by 3D slicer are achieved |