WARNING: the resulting image could have wrong apparent results #2810

jcohenadad · 2020-07-22T18:50:52Z

The following warning is thrown when running sct_register_to_template with the -ref subject flag:

WARNING: the resulting image could have wrong apparent results. You should use an affine transformation as last transformation...

Full output

sct_register_to_template -i t2.nii.gz -s t2_seg.nii.gz -l labels.nii.gz -c t2s -ref subject

--
Spinal Cord Toolbox (git-master-67c679deadd8af0c2eb7445041b0c61628d8bbfa)


Check template files...
  OK: /Users/julien/code/sct/data/PAM50/template/PAM50_t2s.nii.gz
  OK: /Users/julien/code/sct/data/PAM50/template/PAM50_levels.nii.gz
  OK: /Users/julien/code/sct/data/PAM50/template/PAM50_cord.nii.gz

Check parameters:
  Data:                 t2.nii.gz
  Landmarks:            labels.nii.gz
  Segmentation:         t2_seg.nii.gz
  Path template:        /Users/julien/code/sct/data/PAM50
  Remove temp files:    1

Check input labels...

Create temporary folder (/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td)...

Copying input data to tmp folder and convert to nii...

Generate labels from template vertebral labeling
Value = 1 : (70.0, 70.0505921918409, 951.0333382073403) --> ( 70.0, 70.0, 951.0)
Value = 2 : (70.0, 70.30833886308339, 923.9948020349481) --> ( 70.0, 70.0, 924.0)
Value = 3 : (70.0, 70.50121918179356, 889.7335862006682) --> ( 70.0, 71.0, 890.0)
Value = 4 : (70.0, 70.38403779633848, 852.8857673162912) --> ( 70.0, 70.0, 853.0)
Value = 5 : (70.0, 70.30789525112371, 818.2217119405902) --> ( 70.0, 70.0, 818.0)
Value = 6 : (70.0, 70.7014333993707, 786.225032047547) --> ( 70.0, 71.0, 786.0)
Value = 7 : (70.0, 70.80759724209705, 753.9094575256927) --> ( 70.0, 71.0, 754.0)
Value = 8 : (70.0, 70.35343403351428, 715.2977342459287) --> ( 70.0, 70.0, 715.0)
Value = 9 : (70.0, 70.41839798246667, 670.7036147472079) --> ( 70.0, 70.0, 671.0)
Value = 10 : (70.0, 70.41868823000898, 624.8615068668977) --> ( 70.0, 70.0, 625.0)
Value = 11 : (70.0, 70.43996427204287, 577.066096720684) --> ( 70.0, 70.0, 577.0)
Value = 12 : (70.0, 70.4400396432111, 527.0463330029733) --> ( 70.0, 70.0, 527.0)
Value = 13 : (70.0, 70.48122739179244, 475.89622053137083) --> ( 70.0, 70.0, 476.0)
Value = 14 : (70.0, 70.57302038880987, 424.0064011379801) --> ( 70.0, 71.0, 424.0)
Value = 15 : (70.0, 70.64169901804065, 370.32187714089974) --> ( 70.0, 71.0, 370.0)
Value = 16 : (70.0, 70.69621109607577, 316.83818222823635) --> ( 70.0, 71.0, 317.0)
Value = 17 : (70.0, 70.5720703125, 261.3439453125) --> ( 70.0, 71.0, 261.0)
Value = 18 : (70.0, 70.67006390034781, 200.1311979293052) --> ( 70.0, 71.0, 200.0)
Value = 19 : (70.0, 70.47842956120093, 141.44073903002308) --> ( 70.0, 70.0, 141.0)
Value = 20 : (70.0, 70.68440271257172, 92.52008346374544) --> ( 70.0, 71.0, 93.0)

Check if provided labels are available in the template

Binarize segmentation

Done! To view results, type:
fsleyes seg_bin.nii.gz &


Change orientation of input images to RPI...

Remove unused label on template. Keep only label present in the input label image...
sct_label_utils -i template_label_body.nii.gz -o template_label_body.nii.gz -remove-reference /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td/label_projected_rpi.nii.gz # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td
File /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td/label_projected_rpi.nii.gz already exists. Will overwrite it.
File /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td/template_label_body.nii.gz already exists. Will overwrite it.

Create temporary folder (/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn)...

Copying input data to tmp folder and convert to nii...

--
ESTIMATE TRANSFORMATION FOR STEP #0
Registration parameters:
  type ........... label
  algo ........... syn
  slicewise ...... 0
  metric ......... MeanSquares
  iter ........... 10
  smooth ......... 0
  laplacian ...... 0
  shrink ......... 1
  gradStep ....... 0.5
  deformation .... 1x1x0
  init ........... 
  poly ........... 5
  filter_size .... 5
  dof ............ Tx_Ty_Tz_Rx_Ry_Rz_Sz
  smoothWarpXY ... 2
  rot_method ..... pca
Labels src: [[-0.0, 45.720001220703125, -116.84002685546875], [-0.0, 46.220001220703125, -152.84002685546875], [5.0, 45.720001220703125, -116.84002685546875]]
Labels dest: [[7.2192840576171875, -1.7866058349609375, 24.678329467773438], [7.2192840576171875, -0.7866058349609375, -10.321670532226562], [12.219284057617188, -1.7866058349609375, 24.678329467773438]]
Degrees of freedom (dof): Tx_Ty_Tz_Rx_Ry_Rz_Sz
Optimization terminated successfully.
         Current function value: 0.000000
         Iterations: 14
         Function evaluations: 1124
Matrix:
[[ 1.00000000e+00 -1.27249256e-11  1.97608995e-12]
 [ 1.26954035e-11  9.99897996e-01  1.42827999e-02]
 [-2.15763595e-12 -1.42827999e-02  9.99897996e-01]]
Center:
[ 8.88595072 -1.4532725  13.0116628 ]
Translation:
[[  -7.21928406   47.33994039 -141.85168966]]

--
ESTIMATE TRANSFORMATION FOR STEP #1

Apply transformation from previous step

Parse list of warping fields...

Get dimensions of data...
  141 x 141 x 991 x 1

Apply transformation...

Apply transformation and resample to destination space...
/Users/julien/code/sct/bin/isct_antsApplyTransforms -d 3 -i src.nii -o src_reg.nii -t warp_forward_0.txt -r dest_RPI.nii -n 'BSpline[3]' # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
Copy affine matrix from destination space to make sure qform/sform are the same.
WARNING: the resulting image could have wrong apparent results. You should use an affine transformation as last transformation...

Done! To view results, type:
fsleyes dest_RPI.nii src_reg.nii &


Parse list of warping fields...

Get dimensions of data...
  141 x 141 x 991 x 1

Apply transformation...

Apply transformation and resample to destination space...
/Users/julien/code/sct/bin/isct_antsApplyTransforms -d 3 -i src_seg.nii -o src_seg_reg.nii -t warp_forward_0.txt -r dest_seg_RPI.nii -n NearestNeighbor # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
Copy affine matrix from destination space to make sure qform/sform are the same.
WARNING: the resulting image could have wrong apparent results. You should use an affine transformation as last transformation...

Done! To view results, type:
fsleyes dest_seg_RPI.nii src_seg_reg.nii &

Registration parameters:
  type ........... imseg
  algo ........... centermassrot
  slicewise ...... 0
  metric ......... MeanSquares
  iter ........... 10
  smooth ......... 0
  laplacian ...... 0
  shrink ......... 1
  gradStep ....... 0.5
  deformation .... 1x1x0
  init ........... 
  poly ........... 5
  filter_size .... 5
  dof ............ Tx_Ty_Tz_Rx_Ry_Rz
  smoothWarpXY ... 2
  rot_method ..... pcahog

Create temporary folder (/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150939.266722-register-q9cwh82e)...

Copy input data to temp folder...
sct_convert -i src_reg.nii -o /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150939.266722-register-q9cwh82e/src.nii
sct_convert -i src_seg_reg.nii -o /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150939.266722-register-q9cwh82e/src_seg.nii
sct_convert -i dest_RPI.nii -o /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150939.266722-register-q9cwh82e/dest.nii
sct_convert -i dest_seg_RPI.nii -o /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150939.266722-register-q9cwh82e/dest_seg.nii

Get image dimensions of destination image...
  matrix size: 52 x 60 x 55
  voxel size:  1.0mm x 1.0mm x 1.0mm

Split input segmentation...

Split destination segmentation...

Split input image...

Split destination image...
Estimate cord angle for each slice: 100%|████████████████████████| 55/55 [00:00<00:00, 120.47iter/s]
Build 3D deformation field: 100%|████████████████████████████████| 55/55 [00:00<00:00, 100.90iter/s]

Generate warping field...
 --> step1Warp.nii.gz

Generate warping field...
 --> step1InverseWarp.nii.gz

Move warping fields...
cp step1Warp.nii.gz /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
cp step1InverseWarp.nii.gz /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
rm -rf /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150939.266722-register-q9cwh82e

--
ESTIMATE TRANSFORMATION FOR STEP #2

Apply transformation from previous step

Parse list of warping fields...

Get dimensions of data...
  141 x 141 x 991 x 1

Apply transformation...

Apply transformation and resample to destination space...
/Users/julien/code/sct/bin/isct_antsApplyTransforms -d 3 -i src_seg.nii -o src_seg_reg.nii -t warp_forward_1.nii.gz warp_forward_0.txt -r dest_seg_RPI.nii -n NearestNeighbor # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
Copy affine matrix from destination space to make sure qform/sform are the same.

Done! To view results, type:
fsleyes dest_seg_RPI.nii src_seg_reg.nii &

Registration parameters:
  type ........... seg
  algo ........... bsplinesyn
  slicewise ...... 0
  metric ......... MeanSquares
  iter ........... 3
  smooth ......... 1
  laplacian ...... 0
  shrink ......... 1
  gradStep ....... 0.5
  deformation .... 1x1x0
  init ........... 
  poly ........... 5
  filter_size .... 5
  dof ............ Tx_Ty_Tz_Rx_Ry_Rz
  smoothWarpXY ... 2
  rot_method ..... pca
sct_image -i dest_seg_RPI.nii -o dest_seg_RPI_pad.nii -pad 0,0,10 # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn

Estimate transformation
/Users/julien/code/sct/bin/isct_antsRegistration --dimensionality 3 --transform 'bsplinesyn[0.5,1,3]' --metric 'MeanSquares[dest_seg_RPI_pad.nii,src_seg_reg.nii,1,4]' --convergence 3 --shrink-factors 1 --smoothing-sigmas 1mm --restrict-deformation 1x1x0 --output '[step2,src_seg_reg_regStep2.nii]' --interpolation 'BSpline[3]' --verbose 1 # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn

Concatenate transformations...

Parse list of warping fields...

Check file existence...
  OK: dest.nii
  OK: warp_forward_0.txt
  OK: warp_forward_1.nii.gz
  OK: warp_forward_2.nii.gz
/Users/julien/code/sct/bin/isct_ComposeMultiTransform 3 warp_final.nii.gz -R dest.nii warp_forward_2.nii.gz warp_forward_1.nii.gz warp_forward_0.txt # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn

Generate output files...
  File created: warp_src2dest.nii.gz

Parse list of warping fields...

Check file existence...
  OK: src.nii
  OK: warp_inverse_2.nii.gz
  OK: warp_inverse_1.nii.gz
  OK: warp_forward_0.txt
/Users/julien/code/sct/bin/isct_ComposeMultiTransform 3 warp_final.nii.gz -R src.nii -i warp_forward_0.txt warp_inverse_1.nii.gz warp_inverse_2.nii.gz # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn

Generate output files...
  File created: warp_dest2src.nii.gz

Apply transfo source --> dest...

Parse list of warping fields...

Get dimensions of data...
  141 x 141 x 991 x 1

Apply transformation...

Apply transformation and resample to destination space...
/Users/julien/code/sct/bin/isct_antsApplyTransforms -d 3 -i src.nii -o src_reg.nii -t warp_src2dest.nii.gz -r dest.nii -n Linear # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
Copy affine matrix from destination space to make sure qform/sform are the same.

Done! To view results, type:
fsleyes dest.nii src_reg.nii &


Apply transfo dest --> source...

Parse list of warping fields...

Get dimensions of data...
  52 x 60 x 55 x 1

Apply transformation...

Apply transformation and resample to destination space...
/Users/julien/code/sct/bin/isct_antsApplyTransforms -d 3 -i dest.nii -o dest_reg.nii -t warp_dest2src.nii.gz -r src.nii -n Linear # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
Copy affine matrix from destination space to make sure qform/sform are the same.

Done! To view results, type:
fsleyes src.nii dest_reg.nii &


Generate output files...
  File created: template_reg.nii
  File created: data_rpi_reg.nii
  File created: warp_template2data_rpi.nii.gz
  File created: warp_data_rpi2template.nii.gz

Remove temporary files...
rm -rf /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150936.537995-register-bp5l2iyn
sct_apply_transfo -i template.nii -o template2anat.nii.gz -d data.nii -w warp_template2anat.nii.gz -crop 0 # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td
sct_apply_transfo -i data.nii -o anat2template.nii.gz -d template.nii -w warp_anat2template.nii.gz -crop 0 # in /private/var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td

Generate output files...
  WARNING: File warp_template2anat.nii.gz already exists. Deleting it...
  File created: warp_template2anat.nii.gz
  WARNING: File warp_anat2template.nii.gz already exists. Deleting it...
  File created: warp_anat2template.nii.gz
  WARNING: File template2anat.nii.gz already exists. Deleting it...
  File created: template2anat.nii.gz
  WARNING: File anat2template.nii.gz already exists. Deleting it...
  File created: anat2template.nii.gz

Delete temporary files...
rm -rf /var/folders/s8/4qnm5q1n261ch35b5kkclsb00000gn/T/sct-20200722150931.564854-register_to_template-iqdb79td

Finished! Elapsed time: 30s

Done! To view results, type:
fsleyes t2.nii.gz template2anat.nii.gz &


Done! To view results, type:
fsleyes /Users/julien/code/sct/data/PAM50/template/PAM50_t2s.nii.gz anat2template.nii.gz &

Note: this warning is not thrown when using -ref template.

To reproduce

SCT version: v4.3.0
data: sct_testing_data/t2

Also see: http://forum.spinalcordmri.org/t/sct-register-to-template-spurious-warning/441

The text was updated successfully, but these errors were encountered:

jcohenadad · 2020-07-22T19:30:12Z

The culprit is here:
https://github.com/neuropoly/spinalcordtoolbox/blob/544a0267c62137e15953e633a750cabe04912fa6/scripts/sct_apply_transfo.py#L312

The variable isLastAffine is suspicious here (might be some legacy code). I will investigate further and clarify the comments.

Fixes #2810

) Fixes #2810

…inalcordtoolbox#2811) Fixes spinalcordtoolbox#2810

jcohenadad changed the title ~~DRAFT-- WAITING FOR MORE CONTENT-- Unnecessary warning~~ WARNING: the resulting image could have wrong apparent results Jul 22, 2020

jcohenadad self-assigned this Jul 22, 2020

jcohenadad added a commit that referenced this issue Jul 22, 2020

Clarified cropping strategy for sct_apply_transfo; remove warning

3ca7f34

Fixes #2810

jcohenadad mentioned this issue Jul 22, 2020

Clarified cropping strategy for sct_apply_transfo; remove warning #2811

Merged

jcohenadad closed this as completed in #2811 Jul 22, 2020

jcohenadad added a commit that referenced this issue Jul 22, 2020

Clarified cropping strategy for sct_apply_transfo; remove warning (#2811

2081cfa

) Fixes #2810

Drulex pushed a commit to Drulex/spinalcordtoolbox that referenced this issue Sep 30, 2020

Clarified cropping strategy for sct_apply_transfo; remove warning (sp…

6631c30

…inalcordtoolbox#2811) Fixes spinalcordtoolbox#2810

joshuacwnewton added the user requested Raised by user on the SCT forum/email/GitHub. Be sure to notify them when fixed in a release. label Jun 1, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

WARNING: the resulting image could have wrong apparent results #2810

WARNING: the resulting image could have wrong apparent results #2810

jcohenadad commented Jul 22, 2020 •

edited

Loading

jcohenadad commented Jul 22, 2020

WARNING: the resulting image could have wrong apparent results #2810

WARNING: the resulting image could have wrong apparent results #2810

Comments

jcohenadad commented Jul 22, 2020 • edited Loading

To reproduce

jcohenadad commented Jul 22, 2020

jcohenadad commented Jul 22, 2020 •

edited

Loading