Hello all,
I am doing pre-processing for resting-state fMRI volumes, but I have encountered an error and am unsure of how to correct it:
I have previously determined the nonlinear warping transformations from my subjects' anatomical (T1) space to the 2mm MNI template using ANTs. At the end of my pre-processing pipeline, I am trying to transform my subjects' functional volumes (206 TRs of 2sec) into this MNI space by 1) transforming the functional volumes into subject T1 space using mri_vol2vol, and 2) applying the nonlinear warp (from ANTs) to the volumes in subject-T1 space to ultimately have my subjects' functional data in MNI space.
However, I am getting an error while running mri_vol2vol, where it "cannot allocate region." I gather from looking around online that this is a memory issue, but I am not sure how to fix it. Thoughts? It is crucial that I be able to translate my functional data into common MNI space before performing second-level analyses...
I do not encounter this issue when performing this transformation on functional connectivity maps (where I take an ROI defined in MNI space, transform it into T1, and then subject-EPI space, calculate the FC with the rest of the brain, and then invert the process by transforming back into T1, and finally MNI space). How can I get this to work with a larger file?
Thank you in advance for your help,
Ben Mooneyham
mri_vol2vol when used in this way does not manage memory very well. It resamples the time series into a 256^3 volume of floats. Even for a small number of time points, this becomes unmanagebly huge
On 05/21/2014 06:06 PM, Ben Mooneyham wrote:
Hello all,
I am doing pre-processing for resting-state fMRI volumes, but I have encountered an error and am unsure of how to correct it:
I have previously determined the nonlinear warping transformations from my subjects' anatomical (T1) space to the 2mm MNI template using ANTs. At the end of my pre-processing pipeline, I am trying to transform my subjects' functional volumes (206 TRs of 2sec) into this MNI space by 1) transforming the functional volumes into subject T1 space using mri_vol2vol, and 2) applying the nonlinear warp (from ANTs) to the volumes in subject-T1 space to ultimately have my subjects' functional data in MNI space.
However, I am getting an error while running mri_vol2vol, where it "cannot allocate region." I gather from looking around online that this is a memory issue, but I am not sure how to fix it. Thoughts? It is crucial that I be able to translate my functional data into common MNI space before performing second-level analyses...
I do not encounter this issue when performing this transformation on functional connectivity maps (where I take an ROI defined in MNI space, transform it into T1, and then subject-EPI space, calculate the FC with the rest of the brain, and then invert the process by transforming back into T1, and finally MNI space). How can I get this to work with a larger file?
Thank you in advance for your help,
Ben Mooneyham
-- Graduate Student UCSB - Department of Psychological & Brain Sciences Building 429 Room 102 (210) 912-6076 bwmooneyham@gmail.com mailto:bwmooneyham@gmail.com
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Ben, There should be a procsurffast flag called -volproj_lowmem which avoids this mri_vol2vol memory issue. -Louis
On Wed, 21 May 2014, Douglas N Greve wrote:
mri_vol2vol when used in this way does not manage memory very well. It resamples the time series into a 256^3 volume of floats. Even for a small number of time points, this becomes unmanagebly huge
On 05/21/2014 06:06 PM, Ben Mooneyham wrote:
Hello all,
I am doing pre-processing for resting-state fMRI volumes, but I have encountered an error and am unsure of how to correct it:
I have previously determined the nonlinear warping transformations from my subjects' anatomical (T1) space to the 2mm MNI template using ANTs. At the end of my pre-processing pipeline, I am trying to transform my subjects' functional volumes (206 TRs of 2sec) into this MNI space by 1) transforming the functional volumes into subject T1 space using mri_vol2vol, and 2) applying the nonlinear warp (from ANTs) to the volumes in subject-T1 space to ultimately have my subjects' functional data in MNI space.
However, I am getting an error while running mri_vol2vol, where it "cannot allocate region." I gather from looking around online that this is a memory issue, but I am not sure how to fix it. Thoughts? It is crucial that I be able to translate my functional data into common MNI space before performing second-level analyses...
I do not encounter this issue when performing this transformation on functional connectivity maps (where I take an ROI defined in MNI space, transform it into T1, and then subject-EPI space, calculate the FC with the rest of the brain, and then invert the process by transforming back into T1, and finally MNI space). How can I get this to work with a larger file?
Thank you in advance for your help,
Ben Mooneyham
-- Graduate Student UCSB - Department of Psychological & Brain Sciences Building 429 Room 102 (210) 912-6076 bwmooneyham@gmail.com mailto:bwmooneyham@gmail.com
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Ben Mooneyham -
Douglas N Greve -
Louis Nicholas Vinke