Difference between revisions of "2010 Summer Project Week Seg Adapt HNT"
From NAMIC Wiki
Ivan.kolesov (talk | contribs) (Created page with '__NOTOC__ <gallery> Image:PW-MIT2010.png|Projects List Image:Model_3D.png|Prediction for the segmentation of the mandible. </gallery> ==Key…') |
Ivan.kolesov (talk | contribs) |
||
| (3 intermediate revisions by the same user not shown) | |||
| Line 20: | Line 20: | ||
<h3>Approach, Plan</h3> | <h3>Approach, Plan</h3> | ||
| + | * Develop understanding of level set filters in ITK and decide on this or another level set implementation. | ||
| + | * Select an energy for segmentation incorporating the shape prior we have in hand. | ||
| + | * Begin implementing the energy in the framework selected in the first step. | ||
| − | |||
</div> | </div> | ||
| Line 28: | Line 30: | ||
<h3>Progress</h3> | <h3>Progress</h3> | ||
| − | + | *Considered level set filters in ITK and the one used in Robust Statistics Segmentation module | |
| + | *Potential paths for pre-processing patient data sets were examined. | ||
</div> | </div> | ||
Latest revision as of 03:46, 25 June 2010
Home < 2010 Summer Project Week Seg Adapt HNT
Prediction for the segmentation of the mandible.
Key Investigators
- Georgia Tech: Ivan Kolesov and Allen Tannenbaum
- MGH: Gregory Sharp
Objective
We are working on segmenting structures in the head, neck, and thorax that are degraded by artifacts or do not have distinctly visible boundaries in CT imagery. An approach for building a prediction for a structure’s segmentation based on other organs exists. However, an algorithm to use this prediction in the segmentation stage must be developed.
Approach, Plan
- Develop understanding of level set filters in ITK and decide on this or another level set implementation.
- Select an energy for segmentation incorporating the shape prior we have in hand.
- Begin implementing the energy in the framework selected in the first step.
Progress
- Considered level set filters in ITK and the one used in Robust Statistics Segmentation module
- Potential paths for pre-processing patient data sets were examined.
