Title: Geometrical and texture features estimation of lung cancer and TB images using chest X-ray database

Authors: S.A. Patil, V.R. Udupi

Addresses: Textile and Engineering Institute, Ichalkaranji 416115, Maharashtra, India. ' Maratha Mandal's College of Engineering, Belgaum, 59001, Karnataka, India

Abstract: Early detection is the most promising way to enhance a patient|s chance for survival of lung cancer. Detection of disease through image is one of the most challenging tasks in medical image analysis. A computer algorithm for nodule detection in chest radiographs is presented. The algorithm includes four main steps like, image acquisition, image pre-processing, nodule detection, and feature extraction. Total 75 lung cancer and TB images are used during experiment to estimate geometrical and texture features. Manual lung field segmentation with Gray Level Co-occurrence Matrix (GLCM) techniques are used for segmentation and texture features estimation respectively.

Keywords: chest X-rays; active shape modelling; GLCM; gray level co-occurrence matrix; lung field segmentation; geometric features; texture features; lung cancer; TB images; tuberculosis; medical image analysis; chest radiographs; medical imaging; image acquisition; image pre-processing; nodule detection; feature extraction.

DOI: 10.1504/IJBET.2011.040453

International Journal of Biomedical Engineering and Technology, 2011 Vol.6 No.1, pp.58 - 75

Published online: 21 Jan 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article