- Library Home /
- Search Collections /
- Open Collections /
- Browse Collections /
- UBC Theses and Dissertations /
- Three-dimensional multispectral stochastic image segmentation
Open Collections
UBC Theses and Dissertations
UBC Theses and Dissertations
Three-dimensional multispectral stochastic image segmentation Johnston, Brian Gerard
Abstract
Current methods of lesion localization and quantification from magnetic resonance imaging, and other methods of computed tomography fall short of what is needed by clinicians to accurately diagnose pathology and predict clinical outcome. We investigate a method of lesion and tissue segmentation which uses stochastic relaxation techniques in three dimensions, using images from multiple image spectra, to assign partial tissue classification to individual voxels. The algorithm is an extension of the concept of Iterated Conditional Modes first used to restore noisy and corrupted images. Our algorithm requires a minimal learning phase and may incorporate prior organ models to aid in the segmentation. The algorithm is based on local neighbourhoods and can therefore be implemented in parallel to enhance its performance. Parallelism is achieved through the use of a datafiow image processing development package which allows multiple servers to execute in parallel.
Item Metadata
Title |
Three-dimensional multispectral stochastic image segmentation
|
Creator | |
Publisher |
University of British Columbia
|
Date Issued |
1994
|
Description |
Current methods of lesion localization and quantification from magnetic resonance imaging, and
other methods of computed tomography fall short of what is needed by clinicians to accurately
diagnose pathology and predict clinical outcome. We investigate a method of lesion and tissue
segmentation which uses stochastic relaxation techniques in three dimensions, using images from
multiple image spectra, to assign partial tissue classification to individual voxels. The algorithm
is an extension of the concept of Iterated Conditional Modes first used to restore noisy and
corrupted images. Our algorithm requires a minimal learning phase and may incorporate prior
organ models to aid in the segmentation. The algorithm is based on local neighbourhoods and
can therefore be implemented in parallel to enhance its performance. Parallelism is achieved
through the use of a datafiow image processing development package which allows multiple
servers to execute in parallel.
|
Extent |
2790487 bytes
|
Genre | |
Type | |
File Format |
application/pdf
|
Language |
eng
|
Date Available |
2009-02-23
|
Provider |
Vancouver : University of British Columbia Library
|
Rights |
For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.
|
DOI |
10.14288/1.0051371
|
URI | |
Degree | |
Program | |
Affiliation | |
Degree Grantor |
University of British Columbia
|
Graduation Date |
1994-05
|
Campus | |
Scholarly Level |
Graduate
|
Aggregated Source Repository |
DSpace
|
Item Media
Item Citations and Data
Rights
For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.