Loading

Hybridized Algorithms for Medical Image Segmentation
Anusuya Venkatesan1, Latha Parthiban2
1Anusuya Venkatesan, Department of Information Technology, Saveetha University, Chennai, India.
2Latha Parthiban, Department of CSE, Pondicherry University, Pondicherry, India.
Manuscript received on January 26, 2013. | Revised Manuscript received on February 12, 2013. | Manuscript published on February 28, 2013. | PP: 305-307 | Volume-2 Issue-3, February 2013.  | Retrieval Number: C1114022313 /2013©BEIESP

Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Clustering analysisis a unsupervised pattern recognition and groups similar data items into same cluster while dissimilar data item will be moved into different clusters. The purpose of data clustering is to reveal the data patterns and gain some initial insights regarding data distribution. Similarly Image segmentation groups pixels of an image into multiple segments with respect to intensities. This in turn helps to segment objects of interest from the images. In this paper we discuss various segmentation algorithms such as Fuzzyc-means, Maximum Entropy optimized with Particle swarm Optimization to detect abnormalities present in the image. We apply these algorithms on MRI image and Ultra sound images. In order to improve the visibility of ultra sound images, we apply morphological filtering before segmentation. The results section of this paper show the outcome of the algorithms.
Keywords: FCM, Maximum Entropy, PSO, MRI and Ultra sound image.