Feature #566 hadamard motion correction #689

MichaelStritt · 2021-07-01T12:12:50Z

Linked issue

Check out #566

BeatrizPadrela · 2021-07-01T13:49:43Z

Ok, when inducing motion to the first volume, to clearly see if some motion correction was being done, it was not working at all. No motion correction was being applied even if the rp_file was okay.
So I tried to run the motion correction without this specific motion correction for multi-tE, so the normal motion correction approach, only with a single line that defines zigzag=false to multiPLD data, and it worked!!

(difference between volume 0 and volume 8 - before - ASL4D.nii)

(difference between volume 0 and volume 8 - after - rASL4D.nii)

[this was the way I found to show the difference between vol0 and vol8, but if we open the images with fsleyes is really clear that there's no motion after the normal motion correction]

@jan-petr is this enough? I have to show you

jan-petr

Good. Except the HadamardType...

Modules/xASL_module_ASL.m

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

jan-petr

OK.

…otion correction on it

…e motion for all TE of the same PLD and return the original rp files back so the plots work

…ultiTE+singlePLD

…correction of FD{2} for multi-TE and option for multiPLD single TE in line 122

…+multiPLD data

… Hadamard recognition in ASLmodule function

HenkMutsaerts · 2021-07-05T16:52:17Z

Development/xASL_im_Move3DVolume.m

+%
+% FORMAT:       xASL_im_Move3DVolume(pathNifti, pathResult, dim, volNum, numVoxel)
+% 
+% INPUT:        pathNifti (CHAR ARRAY, REQUIRED)


@BeatrizPadrela You need to explain a bit here what the input arguments mean, not only their format.

HenkMutsaerts · 2021-07-05T16:54:04Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

 % Define realignment settings
 tempnii = xASL_io_ReadNifti(InputPath);
 nFrames = double(tempnii.hdr.dim(5));
+if length(x.EchoTime)>1


@BeatrizPadrela This could go wrong for non-multi-TE sequences. The EchoTime can also be (incorrectly) a vector in enhanced DICOM. So we need to add a boolean like if bMultiEcho

HenkMutsaerts · 2021-07-05T16:54:52Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

 % Run SPM
-if nFrames>2 && bSubtraction
+
+%####### Calculating the motion for every first TE of each PLD and


@BeatrizPadrela #### is for languages like Bash, %%%% for Matlab :)

HenkMutsaerts · 2021-07-05T16:57:31Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

+%####### Calculating the motion for every first TE of each PLD and
+% considering it the same for the other TEs from that PLD.
+%
+% if nFrames>2 && bSubtraction && length(x.EchoTime)>1  %Multi TE 


@BeatrizPadrela So this should also start with something like if bMultiEcho. Makes it less prone to error and also easier to read

HenkMutsaerts · 2021-07-05T16:58:06Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

+% considering it the same for the other TEs from that PLD.
+%
+% if nFrames>2 && bSubtraction && length(x.EchoTime)>1  %Multi TE 
+%     uniqueTE=uniquetol(x.EchoTime); %gives the number of unique TEs


@BeatrizPadrela Should we really use uniquetol here or would unique be better? We haven't specified the tolerance here...

HenkMutsaerts · 2021-07-05T16:59:00Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

+%     uniqueTE=uniquetol(x.EchoTime); %gives the number of unique TEs
+%     NumTEs=numel(uniqueTE);
+%     minTE=min(uniqueTE); 
+%     positionMinTE=find(x.EchoTime == minTE); %positions that have the min TE


@BeatrizPadrela At some locations in this script we use TE at other locations EchoTime, though they mean the same right? Would be good to homogenize for readability.

HenkMutsaerts · 2021-07-05T17:09:31Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

+%     if numel(unique(x.Q.Initial_PLD))==1 %multiTE + single PLD
+%         spm_realign(ImInfoFirstTEs,flags,true);
+%         MotionFirstTEs=load(rpfile);
+%         MotionAllTEs=repelem(MotionFirstTEs(:,:),NumTEs,1); %repeats each row NumTEs times


@BeatrizPadrela Is this also possible with repmat? Is a bit more backward compatible.

HenkMutsaerts · 2021-07-05T17:11:38Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m

+%         spm_realign(ImInfoFirstTEs,flags,false);
+%         MotionFirstTEs=load(rpfile);
+%         MotionAllTEs=repelem(MotionFirstTEs(:,:),NumTEs,1); 
+%         save('rp_ASL4D.txt','MotionAllTEs','-ascii')


; missing

HenkMutsaerts · 2021-07-05T17:12:11Z

Modules/SubModule_ASL/xASL_wrp_RealignASL.m


 % Summarize real-world realign parameters into net displacement vector (NDV)
 rp = load(rpfile, '-ascii'); % load the 3 translation and 3 rotation values
 MeanRadius = 50; % typical distance center head to cerebral cortex (Power et al., NeuroImage 2012)


@BeatrizPadrela Apparently, this needs adaptation in some cases. Something for the future :)

MichaelStritt added the feature New feature, enhancement or request label Jul 1, 2021

MichaelStritt assigned BeatrizPadrela Jul 1, 2021

MichaelStritt added this to the Release 1.8.0 milestone Jul 1, 2021

MichaelStritt linked an issue Jul 1, 2021 that may be closed by this pull request

Hadamard_MotionCorrection #566

Closed

MichaelStritt requested a review from jan-petr July 1, 2021 14:01

jan-petr requested changes Jul 2, 2021

View reviewed changes

Modules/xASL_module_ASL.m Outdated Show resolved Hide resolved

Modules/SubModule_ASL/xASL_wrp_RealignASL.m Outdated Show resolved Hide resolved

jan-petr approved these changes Jul 2, 2021

View reviewed changes

jan-petr assigned jan-petr and unassigned BeatrizPadrela Jul 2, 2021

BeatrizPadrela and others added 11 commits July 2, 2021 11:48

#566 xASL_wrp_RealignASL find the first TE for each PLD and perfomr m…

9f2816f

…otion correction on it

#566 xASL_wrp_RealignASL just some adjustments

acd70ab

#566 xASL_wrp_RealignASL changed the motion matrix to include the sam…

9b6e93d

…e motion for all TE of the same PLD and return the original rp files back so the plots work

#566 xASL_wrp_RealignASL motion correction for multiTE+multiPLD and m…

49822b4

…ultiTE+singlePLD

#566 xASL_wrp_RealignASL variable nFrames adapted for multi-TE

5c5dd49

#566 xASL_wrp_RealignASL add an option for plots if ENABLE is on and …

eff4cea

…correction of FD{2} for multi-TE and option for multiPLD single TE in line 122

#566 xASL_wrp_RealignASL add an option for skipping ENABLE if multiTE…

fda8b35

…+multiPLD data

#566 xASL_im_Move3DVolume: Helper function for Beatriz

57455ac

#566 xASL_im_Move3DVolume: Helper function for Beatriz (fix)

2ac6990

#566 changing zigzag to false if multiPLD/multiTE/Hadamard and adding…

7984f1c

… Hadamard recognition in ASLmodule function

#566 Minor fix in wrp_realignASL and module_ASL

af3d8d7

jan-petr force-pushed the feature-#566_HadamardMotionCorrection branch from 352fa81 to af3d8d7 Compare July 2, 2021 09:48

jan-petr merged commit af3d8d7 into develop Jul 2, 2021

jan-petr deleted the feature-#566_HadamardMotionCorrection branch July 2, 2021 09:49

HenkMutsaerts reviewed Jul 5, 2021

View reviewed changes

Feature #566 hadamard motion correction #689

Feature #566 hadamard motion correction #689

Uh oh!

Conversation

MichaelStritt commented Jul 1, 2021

Linked issue

Uh oh!

BeatrizPadrela commented Jul 1, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jan-petr left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

jan-petr left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

BeatrizPadrela commented Jul 1, 2021 •

edited

Loading