An Automated System for Atlas Based Multiple Organ Segmentation of Abdominal CT Images
Sanjay Saxena *
School of Biomedical Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, UP, India.
Neeraj Sharma
School of Biomedical Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, UP, India.
Shiru Sharma
School of Biomedical Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, UP, India.
S. K. Singh
Department of Computer Science and Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, UP, India.
Ashish Verma
Department of Radiodiagnosis and Imaging, Institute of Medical Sciences, Banaras Hindu University, India.
*Author to whom correspondence should be addressed.
Abstract
In this paper, an efficient atlas based approach for multiple abdominal organ segmentation is presented. This automatic segmentation of different organs such as spine, kidneys, liver, aorta, spleen of abdominal image is based on allocation of spine as landmark. In current years several researches has been done for developing automatic segmentation techniques of abdominal CT images however still it is an incredibly challenging task to segment this efficiently and appropriately. This paper proposed a fully automatic system for abdominal image segmentation by marking spine as landmark to extract different organs using a fuzzy based system. The proposed technique uses the fact that multiple organs of abdominal images are situated at a particular distance and in particular range of angles from the spine and spine is the solitary organ which is frequent in the slices of CT image data set. In this paper we focused for the segmentation of liver, kidney, aorta, spine, spleen. This system is evaluated on the data of several patients (152 CT images which consist all such organs) and obtained significant results by comparing the computed results to the boundaries manually traced by experts.
Keywords: Image segmentation, abdominal imaging, medical imaging, atlas, landmark allocation.