American Journal of Applied Sciences

An Adaptive Region Growing Algorithm with Support Vector Machine Classifier for Tuberculosis Cavity Identification

K. Meenakshi Sundaram and C. S. Ravichandran

DOI : 10.3844/ajassp.2013.1616.1628

American Journal of Applied Sciences

Volume 10, Issue 12

Pages 1616-1628

Abstract

The major world health menace is Tuberculosis (TB) that has the effect on one-third of the global population and millions of new infections are occurring every year. The cavities in the upper lung zone are a strong indicator that the disease has developed into a highly contagious state. The study includes 52 Chest X-ray images with Tuberculosis and 43 Chest X-ray images without Tuberculosis. The identification of the TB cavities is mostly conducted by the clinicians by observing the chest radiographs. But the automatic screening has lot of advantages such as substantial reduction in the labor workload of clinicians, enhancing the sensitivity of the test and better precision in diagnosis by increasing the number of images that can be analyzed by the computer. Many researchers have proposed different techniques to improve the performance of automatic screening process. This paper improves the accuracy over the existing technique using the adaptive region growing property and SVM classifier. Initially, pre-processing is carried out for the input image using Gaussian filtering technique to make the image suitable for further processing. The contours of the image will be obtained using region growing technique. The SVM classifier is then used to confirm the suspected TB cavities. The classification will be carried out by the features which the study has taken from the segmented image. The proposed technique is implemented in MATLAB and the performance is compared with the existing technique. From the result, the study has achieved eighty-five percentage accuracy over the existing technique’s seventy eight-percentage accuracy.

Copyright

© 2013 K. Meenakshi Sundaram and C. S. Ravichandran. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.