Difference between revisions of "2014 Summer Project Week:Tumor DCE-MRI Segmentation"
From NAMIC Wiki
| Line 30: | Line 30: | ||
*Segmentation label map further segments tumor-class voxels based on slope ranges of the wash-out curve (persistent, plateau, or washout). | *Segmentation label map further segments tumor-class voxels based on slope ranges of the wash-out curve (persistent, plateau, or washout). | ||
*Interactive graphing of voxel intensity over time at cursor location, and 3D volume rendering of segmentation label map. | *Interactive graphing of voxel intensity over time at cursor location, and 3D volume rendering of segmentation label map. | ||
| + | |||
| + | </div> | ||
| + | <div style="width: 27%; float: left; padding-right: 3%;"> | ||
| + | <h3>References</h3> | ||
| + | *http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Extensions/OpenCAD | ||
</div> | </div> | ||
</div> | </div> | ||
Revision as of 14:18, 23 June 2014
Home < 2014 Summer Project Week:Tumor DCE-MRI Segmentation
Key Investigators
- Vivek Narayan
- Jayender Jagadeesan
Project Description
SegmentCAD module for Slicer (openCAD Extension): Segmentation of Breast Tumors from DCE-MRI images.
Objective
- Add support for DCE-MRI data-sets with more than 4 post-contrast time-points and for 4D Multi-Volume images.
- Implement more sophisticated algorithms for noise reduction
- Add features for intuitive analysis of segmented tumor.
Approach, Plan
- Multiple node input through a dialog box.
- Use established SegmentCAD module framework to test algorithms to optimize segmentation.
- Add GUI features to manually adjust settings and quickly visualize different segmentation outputs.
Progress
- Segmentation for DCE-MRI data-sets with 1 pre-contrast image and 4 post-contrast time-points.
- Segmentation label map further segments tumor-class voxels based on slope ranges of the wash-out curve (persistent, plateau, or washout).
- Interactive graphing of voxel intensity over time at cursor location, and 3D volume rendering of segmentation label map.
