Hi Matt, Michael, and Tim,
I am trying to obtain the BOLD time series registered to native T1 space
for HCP-style dataset acquired at my institution. All the preprocessing
steps were completed with the MPP scripts (prefreesurfer, free surfer, post
frees surfer, Volume fMRI, and Surface fMRI).
So given the files ${fMRIName}_gdc_warp.nii.gz, ${fMRIName}/
MotionMatrices/MAT_####, fMRI2str.nii.gz, I can now use FSL's "convertwarp"
to generate a single deformation field file combining the 3 files, then use
"applywarp" on the original BOLD time series ${fMRIName}_orig.nii.gz in a
single sampling step, right? Please correct me or suggest to
alternative/additions to fully correct the fMRI volumes in native space.
Just a suggestion,having native-space corrected volumetric fMRI data is
useful in clinical cases guiding pre surgical planning, for example,
language/motor localizers. It'll be useful to have a scripts with the MPP
files performing native fMRI-T1-T2 correction and registration.
As always, thanks for your help and guidance.
Osama
*Osama Abdullah*, Ph.D.
MRI Physicist – Core Technology Platforms
New York University Abu Dhabi
Computational Research Building (A2), MRI Lab
Abu Dhabi, UAE (EDT + 8 hours)
Office: +971-2-628-4990
Cell: +971-56-679-3378
On Fri, Jan 12, 2018 at 11:59 PM, Glasser, Matthew <glass...@wustl.edu>
wrote:
> Here are the files you would need:
>
> ???/arc001/${Subject}_3T/RESOURCES/${fMRIName}_preproc/
> ${fMRIName}/${fMRIName}_gdc_warp.nii.gz — Nonlinear gradient warpfield
> ???/${Subject}_3T/RESOURCES/${fMRIName}_preproc/${fMRIName}/MotionMatrices/MAT_####
> — Affine motion matrices from 0000 to #Frames-1
> ???/${Subject}_3T/RESOURCES/${fMRIName}_preproc/${fMRIName}/
> DistortionCorrectionAndEPIToT1wReg_FLIRTBBRAndFreeSurferBBRbased/fMRI2str.nii.gz
> — Distorted EPI to T1w (i.e. Field map correction and registration of EPI
> to T1w)
>
> I don’t know how to put them into a REST call however.
>
> Peace,
>
> Matt.
>
> From: "Hodge, Michael" <hod...@wustl.edu>
> Date: Friday, January 12, 2018 at 8:43 AM
> To: Matt Glasser <glass...@wustl.edu>, Osama Abdullah <o...@nyu.edu>
> Cc: Osama Abdullah <osama.abdul...@utah.edu>, "
> hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org>, "
> at...@nyu.edu" <at...@nyu.edu>, "Brown, Tim" <tbbr...@wustl.edu>
> Subject: RE: [HCP-Users] volumetric BOLD time series in Native Space?
>
>
>
> Hi Osama,
>
>
>
> There’s a wiki page that discusses using the REST interface in
> ConnectomeDB. Additionally, if you search the discussion group archive,
> you’ll also find a number of posts with examples.
>
>
>
> Here’s the wiki page:
>
>
>
> https://wiki.humanconnectome.org/display/PublicData/How+To+
> Access+Subject+Data+via+REST
>
>
>
> Here’s a few posts from the discussion group:
>
>
>
> https://www.mail-archive.com/hcp-users@humanconnectome.org/msg02094.html
>
>
>
> https://www.mail-archive.com/hcp-users@humanconnectome.org/msg05458.html
>
>
>
> https://www.mail-archive.com/hcp-users@humanconnectome.org/msg05429.html
>
>
>
> Regards,
>
>
>
> Mike
>
>
>
>
>
> *From:* Glasser, Matthew
> *Sent:* Friday, January 12, 2018 5:35 AM
> *To:* Osama Abdullah <o...@nyu.edu>
> *Cc:* Osama Abdullah <osama.abdul...@utah.edu>;
> hcp-users@humanconnectome.org; at...@nyu.edu; Hodge, Michael <
> hod...@wustl.edu>; Brown, Tim <tbbr...@wustl.edu>
> *Subject:* Re: [HCP-Users] volumetric BOLD time series in Native Space?
>
>
>
> Perhaps Mike and Tim could help.
>
>
>
> Matt.
>
>
>
> *From: *Osama Abdullah <o...@nyu.edu>
> *Date: *Thursday, January 11, 2018 at 10:48 PM
> *To: *Matt Glasser <glass...@wustl.edu>
> *Cc: *Osama Abdullah <osama.abdul...@utah.edu>, "
> hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org>, "
> at...@nyu.edu" <at...@nyu.edu>
> *Subject: *Re: [HCP-Users] volumetric BOLD time series in Native Space?
>
>
>
> Thanks Matt.
>
>
>
> Yes, would it be possible to get the warpfields (gradient nonlinearity,
> motion correction, distortion correction, and fMRI to T1 native) so that we
> can combine in single step resampling?
>
>
>
> Not sure what’s the REST interface. But I am happy to compare the two
> options.
>
>
>
> All the best.
>
>
>
> Osama
>
>
> On Jan 11, 2018, at 5:42 PM, Glasser, Matthew <glass...@wustl.edu> wrote:
>
> I don’t know that we provide the files to do this yourself in the
> packages. We could consider providing the gradient nonlinearity warpfield,
> the motion correction matrices, the distortion correction warpfield, and
> the fMRI to T1w transformation matrix. Alternatively, you could resample
> the MNI space timeseries with splines back to T1w space. This will induce
> some slight smoothing relative to having done this natively, but I am
> guessing it will not be that different (particularly because of the drift
> in MNI space which leads to an expansion). If you are interested in
> comparing the two options, we could probably help you get the intermediate
> files using the REST interface.
>
>
>
> Peace,
>
>
>
> Matt.
>
>
>
> *From: *<hcp-users-boun...@humanconnectome.org> on behalf of Osama
> Abdullah <osama.abdul...@utah.edu>
> *Date: *Thursday, January 11, 2018 at 12:29 AM
> *To: *"hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org>
> *Cc: *"at...@nyu.edu" <at...@nyu.edu>, "o...@nyu.edu" <o...@nyu.edu>
> *Subject: *[HCP-Users] volumetric BOLD time series in Native Space?
>
>
>
> Dear HCP'ers,
>
>
>
> When using the Minimal Preprocessing Pipeline, seems we don't get
> native-space-registered-to-nativeT1 BOLD time series that
> is gradient-distortion-corrected, motion-corrected, and EPI distortion
> corrected? Am I missing it or does not exist? If it doesn't exist, what
> would be the required modification to the "One Step Spline Resampling" step
> to apply all the corrections *except* the MNI-registraion step?
>
>
>
> Many thanks for your continued superb work.
>
>
>
> Regards.
>
>
>
> Osama
>
>
>
>
>
>
>
> _______________________________________________
> HCP-Users mailing list
> HCP-Users@humanconnectome.org
> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>
>
_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
