Back to Projects List
- Murat Maga (Seattle Children’s Research Institute & University of Washington Dept. of Pediatrics)
- Sara Rolfe (University of Washington Friday Harbor Laboratories)
- Steve Pieper (Isomics)
This project aims to produce a flexible toolset for analysis of shape and form from image data using landmark-based Geometric Morphometric Methods (GMM). This toolset will be capable of handling a broad range of data and file types and will provide support for the full analysis workflow within 3D Slicer.
- Develop a Slicer extension with modules to support each task in the shape analysis workflow from image import through visualization of statistical output.
Approach and Plan
- Provide image import tools that support a broad range of image modalities and filetypes, including non-DICOM image outputs from research microCT scanners.
- Support common preprocessing steps, that may be required before analysis, including downsampling very large files produced by research microCT scanners.
- Develop a module to implement Generalized Procrustes Analysis (GPA) and Principal Component Analysis (PCA) and visualize output statistics.
Progress and Next Steps
- Custom image import tool
- Custom landmark file import tool
- Module to downsize images on import, optimized for speed.
- Module to downsize images on import, optimized for memory usage.
- Module to load large images, crop blank space, and saving ROI.
- Module to perform GPA, PCA, and visualize statistical output
- Added custom layout and display refinements
- Tested GPA module and fixed bugs
- Discussed method for comparing original volume to PCA warped volume. Each will be shown in one of the 3D volume viewers and a button will be added to toggle view.
- Discussed improving 2D vector plots. Plan to implement in VTK and add functionality to Slicer.
Background and References