Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Modified region growing segmentation optimised with genetic algorithm for MRI brain images

Modified region growing segmentation optimised with genetic algorithm for MRI brain images Image segmentation is an important and challenging factor in the medical image processing. This paper describes a new segmentation method that modifies the region growing method. It makes use of texture constraint in addition to intensity constraint to grow the region for the purpose of segmentation. Also it was optimised using genetic algorithm (GA), i.e., it uses GA for choosing intensity and texture constraint. Texture image is obtained from LBP image. At first the film artefact and noise are removed and the image is enhanced using Gaussian filtering and normalisation. Secondly, features are extracted from the image and it is classified using k-NN classifier. In the third phase optimised region growing (ORGW) segmentation was done, if the image is abnormal. Finally the image is compared with the ground truth image and the segmentation accuracy was calculated. Work was carried out for many images. The segmentation accuracy is 75.31% for RGW method and it is 93.57% for ORGW. This automatic detection of brain tumour through MRI can provide the valuable outlook and accuracy of earlier brain tumour detection. Keywords: optimised region growing segmentation; k-NN classifier; magnetic resonance imaging; MRI. Reference to this paper should be made as follows: Viji, http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Enterprise Network Management Inderscience Publishers

Modified region growing segmentation optimised with genetic algorithm for MRI brain images

Loading next page...
 
/lp/inderscience-publishers/modified-region-growing-segmentation-optimised-with-genetic-algorithm-Diue5dqvkD
Publisher
Inderscience Publishers
Copyright
Copyright © 2016 Inderscience Enterprises Ltd.
ISSN
1748-1252
eISSN
1748-1260
DOI
10.1504/IJENM.2016.075173
Publisher site
See Article on Publisher Site

Abstract

Image segmentation is an important and challenging factor in the medical image processing. This paper describes a new segmentation method that modifies the region growing method. It makes use of texture constraint in addition to intensity constraint to grow the region for the purpose of segmentation. Also it was optimised using genetic algorithm (GA), i.e., it uses GA for choosing intensity and texture constraint. Texture image is obtained from LBP image. At first the film artefact and noise are removed and the image is enhanced using Gaussian filtering and normalisation. Secondly, features are extracted from the image and it is classified using k-NN classifier. In the third phase optimised region growing (ORGW) segmentation was done, if the image is abnormal. Finally the image is compared with the ground truth image and the segmentation accuracy was calculated. Work was carried out for many images. The segmentation accuracy is 75.31% for RGW method and it is 93.57% for ORGW. This automatic detection of brain tumour through MRI can provide the valuable outlook and accuracy of earlier brain tumour detection. Keywords: optimised region growing segmentation; k-NN classifier; magnetic resonance imaging; MRI. Reference to this paper should be made as follows: Viji,

Journal

International Journal of Enterprise Network ManagementInderscience Publishers

Published: Jan 1, 2016

There are no references for this article.