An improved AntTree algorithm for MRI brain segmentation

Author

Chenling, Li ; Wenhua, Zeng ; Jiahe, Zhuang

Author_Institution

Software Sch., Xiamen Univ., Xiamen

fYear

2008

fDate

12-14 Dec. 2008

Firstpage

679

Lastpage

683

Abstract

In this paper, an improved method is proposed based on the AntTree algorithm to deal with MRI brain segmentation. This algorithm uses a new tree-structure model to accelerate the calculation of segmenting the brain structure into brain structure, while matter, grey matter, and cerebrospinal fluid. The experimental results indicated that this new approach has made full usage of the pixels information of MRI. Compared with K-means algorithm and FCM algorithm, the results show that the improved AntTree algorithm is characterized by faster, robustness and accurateness.

Keywords

biomedical MRI; brain; fuzzy set theory; image segmentation; medical image processing; pattern clustering; trees (mathematics); AntTree algorithm; K-means algorithm; MRI brain segmentation; cerebrospinal fluid; fuzzy c-means algorithm; magnetic resonance imaging; pixel information; Biomedical imaging; Brain modeling; Clustering algorithms; Feature extraction; Image segmentation; Magnetic resonance imaging; Pathology; Pixel; Robustness; Software algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

IT in Medicine and Education, 2008. ITME 2008. IEEE International Symposium on

Conference_Location

Xiamen

Print_ISBN

978-1-4244-3616-3

Electronic_ISBN

978-1-4244-2511-2

Type

conf

DOI

10.1109/ITME.2008.4743952

Filename

4743952