help
487 Subscribers 
help > Re-run 1.6mm at 2mm, keep all original file?
Showing 1-4 of 4 posts
Feb 5, 2021 05:02 PM | Sarah Kark - UCI
Re-run 1.6mm at 2mm, keep all original file?
Hello,
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
Feb 5, 2021 10:02 PM | Alfonso Nieto-Castanon - Boston University
RE: Re-run 1.6mm at 2mm, keep all original file?
Hi Sarah,
Re-slicing to 2mm isotropic voxels is part of CONN's default preprocessing pipeline (this happens during the normalization step), so your wau* and swau* files should already have 2mm resolution, while your au*, u*, * files will have the same resolution as your original data (1.6mm)
Best
Alfonso
Originally posted by Sarah Kark:
Re-slicing to 2mm isotropic voxels is part of CONN's default preprocessing pipeline (this happens during the normalization step), so your wau* and swau* files should already have 2mm resolution, while your au*, u*, * files will have the same resolution as your original data (1.6mm)
Best
Alfonso
Originally posted by Sarah Kark:
Hello,
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
Feb 8, 2021 12:02 AM | Sarah Kark - UCI
RE: Re-run 1.6mm at 2mm, keep all original file?
Originally posted by Alfonso Nieto-Castanon:
Hi Alfonso,
Thank you very much for your response! For the
7T data, I did not use the defaults, so the input images (e.g., the
first level BETA images) are in 1.6mm isotropic (same resolution as
the original data). I can check and see that using SPM display.
That was intentional, I wanted to have the highest resolution I
could for this analysis, but now I would like to check the results
as if those 1.6mm data at 7T were acquired at 2mm (to test if FC
differences are due to the resolution or the magnet strength). How
might I test that? I imagine re-slicing the 1.6mm cubed BETA images
would be fast but would not quite allow me to answer my question. I
am afraid if I re-run at 2mm it will overwrite everything, but I am
also hesitant to make copies of the raw data to input into a new
CONN project since it's so massive.
Best,
Sarah
Hi
Sarah,
Re-slicing to 2mm isotropic voxels is part of CONN's default preprocessing pipeline (this happens during the normalization step), so your wau* and swau* files should already have 2mm resolution, while your au*, u*, * files will have the same resolution as your original data (1.6mm)
Best
Alfonso
Originally posted by Sarah Kark:
Re-slicing to 2mm isotropic voxels is part of CONN's default preprocessing pipeline (this happens during the normalization step), so your wau* and swau* files should already have 2mm resolution, while your au*, u*, * files will have the same resolution as your original data (1.6mm)
Best
Alfonso
Originally posted by Sarah Kark:
Hello,
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
Feb 14, 2021 08:02 PM | Alfonso Nieto-Castanon - Boston University
RE: Re-run 1.6mm at 2mm, keep all original file?
Hi Sarah,
Sorry there is no straightforward way to do this yet. I would probably suggest to make a backup of your conn_*/results/preprocessing and conn_*/data folders (e.g. simply rename those as conn_*/results/preprocessing_1.6mm and conn_*/data_1.6mm, respectively) then change the 'Analysis space' option in Setup.Options to use a different resolution (this will resample your functional data as part of the Setup step; alternatively you may prefer to resample the functional data directly) and re-run CONN's Setup and Denoising steps with the new 2mm-resampled data (this will fill those data and results/preprocessing folders with the new files). After that, you may directly create in the GUI or though batch scripts new first-level analyses in order to avoid having your original first- and second- level analyses overwritten (e.g. if your original analysis was named SBC_01, then create a new first-level analysis named SBC_01_2mm and run again first- and second-level analysis without risk of overwriting your original SBC_01 results).
After all of this, if you ever need to create new first-level analyses in CONN but again based on the original 1.6mm data, you could do so by first backing up the new conn_*/results/preprocessing folder (e.g. as conn_*/results/preprocessing_2mm), and then renaming the previous backup folder conn_*/results/preprocessing_1.6mm as conn_*/results/preprocessing, and then creating your new first-level analysis as usual in the GUI or through batch scripts.
Hope this helps
Alfonso
Originally posted by Sarah Kark:
Sorry there is no straightforward way to do this yet. I would probably suggest to make a backup of your conn_*/results/preprocessing and conn_*/data folders (e.g. simply rename those as conn_*/results/preprocessing_1.6mm and conn_*/data_1.6mm, respectively) then change the 'Analysis space' option in Setup.Options to use a different resolution (this will resample your functional data as part of the Setup step; alternatively you may prefer to resample the functional data directly) and re-run CONN's Setup and Denoising steps with the new 2mm-resampled data (this will fill those data and results/preprocessing folders with the new files). After that, you may directly create in the GUI or though batch scripts new first-level analyses in order to avoid having your original first- and second- level analyses overwritten (e.g. if your original analysis was named SBC_01, then create a new first-level analysis named SBC_01_2mm and run again first- and second-level analysis without risk of overwriting your original SBC_01 results).
After all of this, if you ever need to create new first-level analyses in CONN but again based on the original 1.6mm data, you could do so by first backing up the new conn_*/results/preprocessing folder (e.g. as conn_*/results/preprocessing_2mm), and then renaming the previous backup folder conn_*/results/preprocessing_1.6mm as conn_*/results/preprocessing, and then creating your new first-level analysis as usual in the GUI or through batch scripts.
Hope this helps
Alfonso
Originally posted by Sarah Kark:
Originally posted by Alfonso
Nieto-Castanon:
Hi Alfonso,
Thank you very much for your response! For the
7T data, I did not use the defaults, so the input images (e.g., the
first level BETA images) are in 1.6mm isotropic (same resolution as
the original data). I can check and see that using SPM display.
That was intentional, I wanted to have the highest resolution I
could for this analysis, but now I would like to check the results
as if those 1.6mm data at 7T were acquired at 2mm (to test if FC
differences are due to the resolution or the magnet strength). How
might I test that? I imagine re-slicing the 1.6mm cubed BETA images
would be fast but would not quite allow me to answer my question. I
am afraid if I re-run at 2mm it will overwrite everything, but I am
also hesitant to make copies of the raw data to input into a new
CONN project since it's so massive.
Best,
Sarah
Hi
Sarah,
Re-slicing to 2mm isotropic voxels is part of CONN's default preprocessing pipeline (this happens during the normalization step), so your wau* and swau* files should already have 2mm resolution, while your au*, u*, * files will have the same resolution as your original data (1.6mm)
Best
Alfonso
Originally posted by Sarah Kark:
Re-slicing to 2mm isotropic voxels is part of CONN's default preprocessing pipeline (this happens during the normalization step), so your wau* and swau* files should already have 2mm resolution, while your au*, u*, * files will have the same resolution as your original data (1.6mm)
Best
Alfonso
Originally posted by Sarah Kark:
Hello,
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah
I completed a project assessing SBC of 7T HCP data (1.6mm). There are some differences between the 7T and 3T data (2mm) so I would like to re-run the 7T data a 2mm to check if it's the resolution of the magnet strength driving the differences.
Is there a way to 'save as' and re-run re-slicing the image to 2mm while maintaining the original 1.6mm files? That is, in the subject data files there are raw files and those pre-pended with wau* and swau*, is there a way to do this without making copies of all of the raw subject data? I have quite a bit of data for 121 subjects so that is a lot of space.
Thank you!
Sarah