Feature Extraction on Survival Predictor data

[dboun]

Describe the bug

The Survival Predictor data confuse Feature Extraction.

To Reproduce

0. Download SurvivalPredictor sample data
1. Load SurvivalPredictor_1.6.1/samples/AAAE/CONVENTIONAL/FLAIR and T1CE into CaPTk.
2. Load SurvivalPredictor_1.6.1/samples/AAAE/SEGMENTATION/SEGMENTATION.nii.gz as ROI
3. Using label 185 in "mask selection" and press "compute + save".
4. Use label 1 and "Selected Image" in "Image Selection". "Computer + save" as [...]/features1.csv
5. Use label 1 and "All images" in "Image Selection". "Computer + save" as [...]/features2.csv

Expected behavior

Step 3: The features for label 185 should be computed. Step 4: X columns of features should be calculated Step 5: 2*X columns of features should be calculated

What happens

Step 3: It says that it finishes, but doesn't save the file specified. For label 25 it saved a file named SubjectID.error that says The ROI for calculation, '25' does not exist in the mask.. Both labels 25 and 185 exist on the ROI, though. It might be because these values are above 9, aka the gui limit? Because it outputs things for labels 1, 2, 4, 7. Step 4: It outputs 4096 columns of features Step 5: It outputs 4096 columns of features

CaPTk Version

1.7.3.beta

Desktop (please complete the following information):

Linux Mint 19.2

[sarthakpati]

It might be because these values are above 9, aka the gui limit

This is exactly the reason why. I am tagging this as a possible future enhancement. This should be possible via the command line, for now.