FALCON is a Python-based software application designed to facilitate PET motion correction, both for head and total-body scans. Our program is built around the fast 'greedy' registration toolkit, which serves as the registration engine. With FALCON, users can enjoy a streamlined experience for implementing motion correction.
When using falconz for 4D total-body PET motion correction, the program is incorrectly including nearly empty moving images for motion correction, despite the fact that NCC should be below the acceptable threshold. This deviates from the expected behavior, where the program should discard such images based on the NCC criteria.
To Reproduce
Steps to reproduce the behavior:
falconz -d <path_to_the_dir> -r deformable
Expected behavior
The program should skip the motion correction for nearly empty moving images if the NCC between the fixed and moving image is less than 0.6 * mean(NCC voxelwise).
Actual behavior
The program is performing motion correction even when the moving image is nearly empty and its NCC with the fixed image does not meet the criteria.
Possible Solution
I am currently investigating this issue. At this moment, a fix or workaround has not been identified. Any contributions or insights into this issue are highly appreciated.
Describe the bug
When using
falconz
for 4D total-body PET motion correction, the program is incorrectly including nearly empty moving images for motion correction, despite the fact that NCC should be below the acceptable threshold. This deviates from the expected behavior, where the program should discard such images based on the NCC criteria.To Reproduce
Steps to reproduce the behavior:
Expected behavior
The program should skip the motion correction for nearly empty moving images if the NCC between the fixed and moving image is less than 0.6 * mean(NCC voxelwise).
Actual behavior
The program is performing motion correction even when the moving image is nearly empty and its NCC with the fixed image does not meet the criteria.
Possible Solution
I am currently investigating this issue. At this moment, a fix or workaround has not been identified. Any contributions or insights into this issue are highly appreciated.