Brain image segmentation is an active research in computer image analysis. The challenge lies in the fact that the brain anatomy is not identical for all normal subjects let alone subjects with abnormal tissue. / In this thesis, we explain the research we have implemented to extract the brain from T1-weighted MRI images, and then segment the brain into the three prominent compartments (i.e. the cerebellum and the two hemispheres of the cerebrum). The brain extraction is implemented using morphological operations after thresholding. The brain segmentation, however, is implemented in two separate steps. The first step segments the two hemispheres by approximating the midsagittal surface using mainly Radon transform. The second step segments the cerebellum using an atlas-based contour as an initial contour for the gradient vector flow active contour algorithm. / Validation tests have been performed for the brain extraction and cerebellum segmentation methods.
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.97949 |
| Date | January 2006 |
| Creators | Ezzadeen, Hani. |
| Publisher | McGill University |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Format | application/pdf |
| Coverage | Master of Science (School of Computer Science.) |
| Rights | © Hani Ezzadeen, 2006 |
| Relation | alephsysno: 002479784, proquestno: AAIMR24664, Theses scanned by UMI/ProQuest. |
Page generated in 0.002 seconds