Skip to Main content Skip to Navigation
Conference papers

Segmentation of Small Animal Computed Tomography Images Using Original CT Values and Converted Grayscale Values

Abstract : Medical image segmentation is the foundation of normal and diseased tissue 3D visualization, operation simulation and visual operation. In this paper, we comprehensively use two values that represent the same object (body tissue), to segment by the same algorithm implementation. The original CT images are downloaded from the web source, the dog rib tumor CT scan image by GE medical systems, all the experiment of dataset of 312 thoracic CT scans. The core of the segmentation is k-means clustering algorithm. The segmentation process consisted of two phases: (1) convert CT value to JPG gray value or not use the original CT value as the data sets for clustering; (2) segmentation bone tissue using the new k-means clustering algorithm program which is implemented with MATLAB 2012a programming language and for two-dimensional data matrix directly. The experiment produced strikingly different results. These results may be indicating that not only the segmentation algorithm of CT image is important, but also the data for segmentation is important too.
Document type :
Conference papers
Complete list of metadata

Cited literature [10 references]  Display  Hide  Download

https://hal.inria.fr/hal-01220970
Contributor : Hal Ifip <>
Submitted on : Tuesday, October 27, 2015 - 10:55:38 AM
Last modification on : Wednesday, January 17, 2018 - 10:43:35 AM
Long-term archiving on: : Thursday, January 28, 2016 - 10:40:26 AM

File

978-3-642-54344-9_54_Chapter.p...
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Guoqiang Ma, Naixiang Li, Xiaojuan Wang. Segmentation of Small Animal Computed Tomography Images Using Original CT Values and Converted Grayscale Values. 7th International Conference on Computer and Computing Technologies in Agriculture (CCTA), Sep 2013, Beijing, China. pp.470-477, ⟨10.1007/978-3-642-54344-9_54⟩. ⟨hal-01220970⟩

Share

Metrics

Record views

1000

Files downloads

320