Research On Image Segmentation Algorithm And Its Application In Diagnosis Of Solitary Pulmonary Nodules

The research on algorithms for medical image segmentation is a multi-disciplinary subject. Segmentation methods are widely used in diagnostic medicine, surgery planning and simulating, plastic and artificial limb surgery,radiotherapy planning, and anatomy teaching. Images segmentation is an important process of image analysis.So study on the segmentation on medical images has important significance to science and worthiness in practical applications.The Computer Aided Diagnosis of SPNs problem is studied in this thesis. The main aspects of research include:First, four segmentation algorithms for medical images are analyzed in this dissertation. And these segmentation algorithms are realized. Then the advantage and disadvantage of each method are presented.Second, a new hybrid algorithm, named as remembered genetic simulated annealing algorithm, is proposed. This algorithm is designed by combing genetic simulated annealing algorithm and maximum entropy theory.Third, a threshold segmentation method is applied to segment lung CT images automatically. This process consists of following steps: filtering the images, binarizing the images by creating the threshold values, using the morphology method on the binary image, using the edge detect method on the binary image, using the 8-neighborhood label algorithm to label the connected components of the binary image,using the rules to extract the pulmonary parenchyma,mending the pulmonary parenchyma if necessary, extracting the regions of the interest(ROI).Fourth, a system is implemented for labeling the SPNs regions in pulmonary CT images. It can help to get the training dataset from medical experts.Fifth, a Computer-Aided Diagnosis (CAD) system for SPNs is achieved.At the end of this dissertation, experiments are carried out to evaluate the performance of the methods above. The experimental results strongly supported the feasibility of these methods. And they can solve the pulmonary CAD problem in a certain extent。...