Lung nodule segmentation in computed tomography image

Abstract

The key process to detect the Lung cancer is the critical observation on the CT scan images where the visual inspection carefully analysis the geometrical features of blob shaped objects found inside the lung region. This traditional approach becomes tedious for the clinical experts when analyzing large amount of CT images and courses to introduce error in the prediction of the cancer nodules. Further the detection of blobs touching the inner wall of the chest is very challenging task and most of the proposed methods are time consuming due to the complex architecture of the algorithms. Therefore, in this paper a computer vision solution is proposed to automatically localize the lung nodules and to overcome the difficulties found in the earlier methods. Experimentation is conducted on CT scan data and shows promising results in the localization of the lung nodules.