Research on computer-aided diagnosis of lung nodule

Abstract

With the development of Computer in recent years, computer-Aided Diagnosis becomes a new technology applied in image processing techniques. For analyzing lung CT image as the research object, pulmonary nodules of computer automatic detection as the goal, the main region are in the segmentation of lung parenchyma, perceptual ROI extraction, feature extraction. The article presents two ways of image-preprocessing and adaptive threshold binarization, after getting the preliminary pulmonary parenchyma, taking the method of removal of trachea and bronchia in region-growing and edge patch with rolling ball, resulting the final segmentation of pulmonary areas.