Multilevel Dapple Based Context Analysis In Pulmonary Nodule
Lung nodule or lung lobe is a small masses of tissue in the lung as quite common. Now adays, the image processing mechanisms are used in number of medical profession for improving detection of lung cancer. The identification of the lobar fissures in CT images are very difficult even for experienced surgeons because of its variable shape and their appearance along with low contrast and high noise with it. Existing works on lung nodule were carried out just to detect the presence or absence in
... ce or absence in nodule using SVM with MR-8 classifier. In this paper presents a novel recognition method of pulmonary nodule but in this the detection of intersection and overlapping in a pulmonary nodule is performed based on the concept of multilevel dapple which partitions the image into an orderless collection of small patches. The classification is carried out using Neural network with MR-16 filters which provides a clear description of patches. This technique is a supervised learning system. In this paper also describes a three steps to extract features are; lung segmentation with adaptive fissure to segments the lobe then detection level carried out by canny edge to detect accurate location then the classifier used as MR-16.It is to reduce redundancy of image data in order to be able to store or transmit data in an efficient form. The main contribution of proposed is to find lobar fissures with accurate detection rate.