Automatic Segmentation of Liver and Lesions using Hybrid Techniques

  • Anusha Linda Kostka. J. E, Dr. S. Vinila Jinny

Abstract

In the recent years, it is observed that liver diseases have caused innumerable deaths to the world. Hence it becomes essential to identify the causes of liver diseases and to take the necessary precautions against it nowadays. This can be done by segmenting the liver from the other organs of the human body and identifying the tumors and hepatic lesions. Among the imaging instruments, the Computed Tomography images can be used for these purposes. It is found that the Computed Tomography images have been utilized more for the diagnosis and detection of the abnormalities in the liver, surgery as well as transplantation. The proposed framework initiates with the preprocessing stage, in which a new filter the bitonic is compared with other noisy images to the anisotropic diffusion filter, Gaussian filter and the median filter. In the first phase, the seed point selection method is followed by the extraction of the liver contour which can be further enhanced using the connected component and the cavity filling method. During the second phase the segmentation is carried on with the watershed algorithm to improve the accuracy of the hepatic liver lesions. The performance results reveal that the proposed work is fast and effective and the accuracy of the bitonic filter provides better visual rates than the other filters.

Published
2020-06-01
How to Cite
Anusha Linda Kostka. J. E, Dr. S. Vinila Jinny. (2020). Automatic Segmentation of Liver and Lesions using Hybrid Techniques. International Journal of Advanced Science and Technology, 29(10s), 8091-8096. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/24259
Section
Articles