Algorithm:UNC:DTI Tract Statistics Workflow

1. Load diffusion weighted imaging with seven scans in Basser gradient scheme or using .nrrd file with gradient metadata
2. Estimate tensors and compute derived tensors measures
3. Save FA image to identify ROIs
4. Load FA image in InsightSNAP
5. Draw source and target ROI for tracking
6. Load ROIs in FiberTracking and compute fiber tracts
7. Save resulting fiber tracts
8. Load resulting fiber tracts in FiberViewer
9. Add image for background overlay (optional)
10. Cluster fiber tracts to clean results
    1. Length filtering
    2. Center of gravity based Hierarchical Agglomerative Clustering (HAC): Useful for removing outliers
    3. Mean distance based Hierarchical Agglomerative Clustering (HAC): Useful for removing outliers
    4. Hausdorff distance based Hierarchical Agglomerative Clustering (HAC): Useful for seperating sections of bundles with small deviations at one end
    5. Normalized cut clustering based on mean distance: Research clustering method
11. Manual tract editing
    1. Cutting fibers with plane
    2. Resampling fibers
12. Tract based statistics
    1. Averaged derived properties as function of arc-length
    2. Average tensor as function of arc-length