An improved Bayesian Network Model Based Image Segmentation in detection of lung cancer

Abstract

User assisted segmentation of lung parenchyma pathology bearing regions becomes difficult with an enormous volume of images. A novel technique using Bayesian Network Model Based (BNMB) Image Segmentation, which is a probabilistic graphical model for segmentation of lung tissues from the X-ray Computed Tomography (CT) images of chest, is proposed. Goal of this work is to present an automated approach to segmentation of lung parenchyma from the rest of chest CT image. This is implemented with help of a probabilistic graph construction from an over-segmentation of the image to represent the relations between the super pixel regions and edge segments. Using an iterative procedure based on the probabilistic model, we identify regions and then these regions are merged. The BNMB is evaluated on many CT image databases and the result shows higher accuracy and efficiency for both segmenting the CT image of lung and also extraction of the Region Of Interest (ROI) from affected CT image.