Application Of Selective Region Growing Algorithm In Lung Nodule Segmentation

Abstract

An effective methodology is proposed for early detection of lung cancer using computed tomography. A detailed literature survey and real time happenings led to capture the abnormal nodules. Initially, gaussian filter was used in pre-processing to remove noise. For segmenting lung parenchyma, an adaptive intensity thresholding method was used. To detect nodules, morphological operation as well as selective region growing algorithm was used. The abnormal nodules were detected by filtering the segmented image based on area.