Combine AntsMotioncorr realigment with transfromation to MNI space pasing through anatomical space

[vinferrer]

Hi,

I was following the example for fMRI data in https://stnava.github.io/fMRIANTs/ and I was wondering how to combine the last transform (${nm}_diffCollapsedWarp.nii.gz) that combines all the motion corrections with my transformations to the MNI space. I use an intermediate step passsing throught the anatomical space. So I wan't to combine ${nm}_diffCollapsedWarp.nii.gz, func2anatGenericaffine.mat, anat2MNIGenericAffine.mat ,MNIwrapping.nii.gz

[stnava]

That example shows the necessary steps to do what you want. Just repeat with another mapping passed via -t.

On Thu, Jul 26, 2018, 06:12 sento4000 notifications@github.com wrote:

Hi,

I was following the example for fMRI data in https://stnava.github.io/fMRIANTs/ and I was wondering how to combine the last transform (${nm}_diffCollapsedWarp.nii.gz) that combines all the motion corrections with my transformations to the MNI space. I use an intermediate step passsing throught the anatomical space. So I wan't to combine ${nm}_diffCollapsedWarp.nii.gz, func2anatGenericaffine.mat, anat2MNIGenericAffine.mat ,MNIwrapping.nii.gz

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ANTsX/ANTs/issues/615, or mute the thread https://github.com/notifications/unsubscribe-auth/AATyfu1-HBnHFI3IxET31tUwVG6Gh1MMks5uKZX3gaJpZM4Vhi8w .

[vinferrer]

now I get it, however there is something i don't understand. you apply motioncorrection with this command: tx=SyN[0.1,3,0.0] # critical parameters (though others matter too) antsRegistration --dimensionality 4 -f 1 -r ${nm}Warp.nii.gz \ --output [${nm}_,${nm}Warped.nii.gz] \ --interpolation Linear --use-histogram-matching 1 \ --winsorize-image-intensities [0.005,0.995] --transform $tx \ --metric meansquares[${fxd},$fmri,1] \ --convergence [15x2,1e-6,4] --shrink-factors 2x1 \ --smoothing-sigmas 1x0vox --restrict-deformation 1x1x1x0

and after computing and combining the other transformations you use this command to apply them to the original bold data: antsApplyTransforms -d 4 -o ${nm}_bold2template.nii.gz \ -t ${nm}_diff4DCollapsedWarp.nii.gz -t ${nm}_0Warp.nii.gz \ -r data/template_replicated.nii.gz -i ${nm}Warped.nii.gz

but ${nm}Warped.nii.gz isn't the original data, is the output of the first command (the replicated templated, warped to the non corrected motion image?). Isn't the original data fmri=data/bold.nii.gz, why isn't this the input?

[vinferrer]

sry that command maps the the time series to the fixed space but anyway i still not get why ${nm}Warped.nii.gz is the input when you apply the transformation to the original bold data.

[stnava]

You can probably figure out another way to do it and feel free to modify the example and share with us

On Sun, Jul 29, 2018, 12:31 sento4000 notifications@github.com wrote:

sry that command maps the the time series to the fixed space but anyway i still not get why ${nm}Warped.nii.gz is the input when you apply the transformation to the original bold data.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ANTsX/ANTs/issues/615#issuecomment-408689244, or mute the thread https://github.com/notifications/unsubscribe-auth/AATyfkMudP5igAZi-_CISZb-NR9mxaxuks5uLeNTgaJpZM4Vhi8w .

[cookpa]

I made a PR

https://github.com/stnava/fMRIANTs/pull/4

This applies the warp to the original bold data.