Here is an example of a skill that can be used to automate the segmentation of multiple datasets using different segmentation tools:
---
name: my-segment-skill
description: Segmentation workflow for brain MRI datasets.
metadata:
skill-author: Talk2View
---
# My Segment Skill
## Datasets
Process these datasets from Slicer sample data in order:
- MR Brain Tumor 1
- MR Brain Tumor 2
## Segments
For each dataset, create these segments:
| Name | Color |
| ---------- | ----- |
| Tumor | red |
| Ventricles | blue |
## How to segment
### Tumor
Adaptive Brush with MRI T1+GD enhancing tumor preset, paint at RAS [-5, 27, 28] with 20 mm radius
### Ventricles
First, jump to RAS [18, 21, 1] in all slice views, which is close to where the ventricles are.
Then use Adaptive Brush with simple threshold algorithm, radius of 7 mm, threshold zone of 30%.
## Workflow
Ensure that the scene is empty before you create a new segmentation.
For each dataset:
1. Download the sample data
2. Create a segmentation with the first segment from the table
3. Activate the tool with the specified preset or algorithm, then apply any extras
4. Show the segmentation in 3D, and if the segment has already been painted and is not empty, jump to the segment in all slice views
5. Tell the user they can now refine the segment, and wait until they are done
6. When the user is finished, add the next segment and repeat from step 3
7. When all segments are done, report the volume of each segment,
save the scene, and clear the scene before
moving to the next dataset
You can use the skills editor inside the Talk2View Slicer extension to create and edit new and existing skills.
If someone has any ideas for workflows to test we would be very happy to hear about them.