We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Hierarchical fuzzy segmentation of brain MR images.
- Authors
Kwon, M. J.; Han, Y. J.; Shin, I. H.; Park, H. W.
- Abstract
In brain magnetic resonance (MR) images, image segmentation and 3D visualization are very useful tools for the diagnosis of abnormalities. Segmentation of white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is the basic process for 3D visualization of brain MR images. Of the many algorithms, the fuzzy c-means (FCM) technique has been widely used for segmentation of brain MR images. However, the FCM technique does not yield sufficient results under radio frequency (RF) nonuniformity. We propose a hierarchical FCM (HFCM), which provides good segmentation results under RF nonuniformity and does not require any parameter setting. We also generate Talairach templates of the brain that are deformed to 3D brain MR images. Using the deformed templates, only the cerebrum region is extracted from the 3D brain MR images. Then, the proposed HFCM partitions the cerebrum region into WM, GM, and CSF. © 2003 Wiley Periodicals, Inc. Int J Imaging Syst Technol 13, 115–125, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.10035
- Subjects
BRAIN abnormalities; DIAGNOSTIC imaging; RESONANCE; MEDICAL imaging systems; MEDICAL equipment; MEDICAL research
- Publication
International Journal of Imaging Systems & Technology, 2003, Vol 13, Issue 2, p115
- ISSN
0899-9457
- Publication type
Article
- DOI
10.1002/ima.10035