vfonov
commented
1 year ago Hello,
sorry for the lack of documentation, you are the first to try to use these scripts in the last 10 years.
Briefly, the phantom scans need to be preprocessed first, and then phantom_distortion_measure_v2.pl script also needs a few extra arguments to function properly.
It should be roughly like following, you need to provide also phantom_model.mnc - "ideal" phantom, phantom_model_mask.mnc - ideal phantom mask, phantom_bricks_map.mnc - map of lego bricks inside of the ideal phantom.
# preprocess scans separately
lego_core_extract.pl raw_scan1.mnc preproc_scan1.mnc preproc_scan1_mask.mnc --grayscale --nuc --denoise
lego_core_extract.pl raw_scan2.mnc preproc_scan2.mnc preproc_scan2_mask.mnc --grayscale --nuc --denoise
....
lego_core_extract.pl raw_scanN.mnc preproc_scanN.mnc preproc_scanN_mask.mnc --grayscale --nuc --denoise
# estimate distortion field with mean square approximation using all scans
phantom_distortion_measure_v2.pl \
--model phantom_model.mnc --mask phantom_model_mask.mnc --bricks phantom_bricks_map.mnc \
--work_dir (temprorary work dir) \
--order 5 --elastix \
preproc_scan1.mnc preproc_scan1_mask.mnc \
preproc_scan2.mnc preproc_scan2_mask.mnc \
.... \
preproc_scanN.mnc preproc_scanN_mask.mnc \
output_par.txt output_distcorr.xfm output_stats.csv --out-roi output_roi_mask.mnc
Hi,
I'm trying to use the
phantom_distortion_measure_v2.pltool to map geometric distortions in phantom scans, but I'm unfortunately running into some problems. I'm running minc on a Mac with the M1 chip. Initially I had some trouble with the installation so I installed everything in a docker container running ubuntu where it seems to work fine. However, I face the same error in the docker as I did when I tried to run it natively on my Mac. Below is the output when I try to run it where I have two near identical scans as a first simple test, both using the same mask. The last output which says "DIED" doesn't give me very much help to figure out what the issue is. Do you have any suggestions for what I can do to investigate this further or have you seen anything like this before?