Testing
Regression testing is performed using 3D Slicer's built in Unittest framework. No additional installations are required.
To run tests, enable Developer Mode in 3D Slicer.
Go to Edit -> Application Settings -> Developer and toggle on Enable Developer Mode
Once this setting is enabled, you will see a collapsible named Reload & Test in each widget. Click Reload and Test to initialize run the tests for each Module. Detailed test results are displayed in the Python Interactor window.
Each test uses 3 XtremeCT II images with known erosions. Files for testing are located in the TestFiles folder in the repository. In addition to the specified tests, each module has failure tests. These will generate a series of errors, which can be ignored if the test has not failed.
Automatic Contour
Generates the contour mask for each image and compares it to a mask created in IPL.
- The number of voxels that are different must be less than 2% of the total number of positive voxels in comparison image
Cortical Break Detection
Generates a cortical break mask and seed points for each image. The output mask is compared to a pre-generated mask, while the seed list is compared to search for a manually selected seed point.
- The number of voxels that are different between the masks must be less than 0.5% of the total number of positive voxels in comparison image
- There must be less than 10 additional seed points detected
- One seed point must be within 1 voxel of the true seed point in all dimensions
Erosion Volume
Generates the erosion segmentation and calculates statistics for each image. The output segmentation and statistics are compared to a manually performed analysis.
- The number of voxels that are different between the segmentations must be less than 2% of the total number of positive voxels in comparison segment
- The difference in volume and surface area must be less than 0.5% of the comparison values
File Converter
Converts an .AIM file to .MHA and compares it to a conversion script output.
- The images must match exactly
