Comparisons on segmentation of brain MR image - Details

Author：

Chunlan, Yang (Chunlan, Yang.) | Shuicai, Wu (Shuicai, Wu.) (Scholars：吴水才) | Yanping, Bai (Yanping, Bai.) | Hongjian, Gao (Hongjian, Gao.)

Indexed by：

EI Scopus

Abstract：

Image　segmentation　is　a　focused　issue　in　image　processing.　Especially,　brain　segmentation　is　a　key　problem　in　neuroscience.　In　this　study,　our　aim　is　to　segment　the　real　MR　image　into　gray　matter,　white　matter　and　cerebrospinal　fluid.　Several　methods　were　compared.　However,　traditional　methods　such　as　fuzzy　c-means,　mixture　Gaussian　model　can＇t　achieve　a　satisfied　result　successfully.　Markov　random　field　(MRF)　model　is　used　and　the　experimental　results　show　that　MRF　method　is　robust　to　noise　which　can　achieves　a　perfect　segmentation.　©2009　IEEE.

Keyword：

Magnetic resonance imaging Mixtures Image segmentation Fuzzy systems Gaussian distribution Cerebrospinal fluid Magnetorheological fluids Markov processes

Author Community：

[ 1 ] [Chunlan, Yang]College of Life Science and Bioengineering, Beijing University of Technology, China
[ 2 ] [Shuicai, Wu]College of Life Science and Bioengineering, Beijing University of Technology, China
[ 3 ] [Yanping, Bai]College of Life Science and Bioengineering, Beijing University of Technology, China
[ 4 ] [Hongjian, Gao]College of Life Science and Bioengineering, Beijing University of Technology, China