Lungs segmentation using wavelet packet frames (P-0099) (MFN 4675)

Abstract

Computer-Aided Diagnosis of lung nodules offers a more exact technique for nodule identification which prompts dependable analysis of lung tumor. The segmentation of lungs is a first phase for the detection of nodules. In this project, the segmentation of lungs is successfully done by wavelet packet frames. The proposed technique chooses the ideal wavelet representation that is a gathering of wavelet packet frames. The frames are in this way utilized for grouping of coefficients utilizing k-means, density based and mean shift clustering which prompts the segmented lung area. Besides, the proposed system is completely computerized and is fit for sectioning lung tissue in various cuts with no change in parameters or manual mediation.