An Improved K-Means Clustering Method for Liver Segmentation

Abstract

Determining liver accurately from CT scans is the primary and crucial step for an automated liver segmentation. The knowledge of the liver structure, liver surface and liver volume is required for liver segmentation. The boundaries of the various organs are nor clearly visible as a result of complex structure of the human body. This paper proposes an ant colony based k-means method which reduces the initial clusters problem of k-means clustering method. In this proposed method level set methods have also been used to improve the contours of the liver region. The paper aims in comparing the traditional k-means method and improved kmeans method using ant colony optimization on the basis of geometric accuracy and elapsed time. Experimental results obtained on 15 CT scan images show that the proposed approach obtained better segmentation results than the already existing one. The results provided an increase in the geometric accuracy and a decrease in elapsed time which clearly show that the results are better than those obtained with existing technique.

Diagnosis of Liver Disorder Using Clustering Technique

“…In the year of 2015 Dr.S.S.Dhenakaran, N.Roobini [1], have proposed a Comparative Analysis of K-means Algorithm in Disease Prediction to analyze the various whereas Harsha Pakhale, Deepak Kumar Xaxa [2], have surveyed many data mining techniques that have used for classification of liver patients. The authors [3], used the improved the K-means clustering for liver segmentation by making it less dependent on the initial parameters such as randomly chosen initial cluster centers and hence more stable.…”