If the goal is to normalize functionals into MNI space, while taking into account the subject's anatomy as well as warping due to EPI distortion in the functional images, would the following steps seem reasonable:
1. coregister one functional to structural (T1, in my case)
2. realign and reslice functional images to coregistered functional image
3. segment the structural image
4. use normalize:estimate with the mean functional image (from step 2) as the source and EPI template and modulated spatially normalised tissue maps of grey and white
matter (mwc1X.img and mwc2X.img) (from step 3) as templates to estimate warps to use to write normalize functional images
Thank-you for your (everyone's) advice re: this topic,
Julie
Julie E. McEntee, M.A., C.C.R.P.
Senior Research Program Coordinator
Department of Psychiatry- Neuroimaging
Johns Hopkins University School of Medicine
600 N. Wolfe St./ Phipps 300
Baltimore, MD 21287
Phone: 410-502-0468
Fax: 410-614-3676
----- Original Message -----
From: Douglas Burman <[log in to unmask]>
Date: Wednesday, July 9, 2008 9:35 pm
Subject: Re: [SPM] individual functional and anatomical superimposition
To: [log in to unmask]
> Tao and Haibo,If you want to save time while preserving resolution,
> you can change the order slightly -- coregister one functional image
> to the structural, then realign and reslice all other functional
> images to this coregistered functional image before smoothing. The
> results are actually quite nice, but not in standard
> coordinates.Doug==============Original message text===============
> On Wed, 09 Jul 2008 8:06:01 pm CDT xuhaibo wrote:
>
>
> Tao,
>
> Without normalisation, just do realigning, coregistering the
> structural to one functional image(rather than vice versa, the former
> saving time, the latter saving resolution) and smoothing as mentioned
> by Doug, then superimposing the contrast maps on the subject's
> anatomical structure. But it does not work very well compared to
> superimposition after normalising both individual structural and
> functional images.
>
> Haibo
>
>
>
> Date: Wed, 9 Jul 2008 18:16:47 +0200From: [log in to unmask] CNRS.FRSubject:
> Re: [SPM] individual functional and anatomical superimpositionTo: [log in to unmask]
>
>
>
> Doug,
> In this case, what is the interest to do the step without
> normalisation ?
> Tao
>
>
> ----- Original Message -----
> From: Douglas Burman
> To: [log in to unmask]
> Sent: Wednesday, July 09, 2008 6:05 PM
> Subject: Re: [SPM] individual functional and anatomical superimposition
>
> Tao,
> That's right. If you additionally want to identify the exact region
> based upon atlas coordinates, you should run both types of analyses in
> parallel -- i.e., with and without normalization.
> Doug==============Original message text===============On Wed, 09 Jul
> 2008 10:44:47 am CDT "jiang" wrote:Doug, Yes, your suggestion helps
> very much the comprehension of what I have done. thank you!But another
> question raised, in this case (without normalisation), does it mean
> that I cannot use the common atlas coordinates to identify the exact
> region where are my activations ?Tao ----- Original Message -----
> From: Douglas Burman To: [log in to unmask] Sent: Wednesday, July 09,
> 2008 3:15 PMSubject: Re: [SPM] noramlisationTao,If you want to
> superimpose the subject's functional activation on his own anatomical
> (structural) image, run a set of analyses on images that were
> realigned and smoothed without normalization. The functional
> activation can then be superimposed on the structural image.To avoid
> loss of resolution for the structural image, it works best to start by
> coregistering one functional image to the structural (rather than vice
> versa), then realign all the functional images to this coregistered
> image. For this analysis, realign and reslice, then smooth the
> resulting images before running the analysis. Hope this
> helps.Doug==============Original message text===============On Wed, 09
> Jul 2008 1:51:04 am CDT jiang wrote:Dear Haibo, Thank you for the
> extension installation complementary information.About normalisation,
> I have been instructed by two different methods, one like written in
> the manual of SPM, s egmentation--coregistre anatomical and mean
> realigned functional images, use the segmentation parameter to
> normalise functional and anatomical ones ; 2nd method : nomalise mean
> functional to EPI t emplet, and then use the estimated parameter to
> normalise the coregiestered anatomical image. Which is better to
> superimpose the functional image and anatomical one ?Another question
> how to do to superimpose one's functional image on his own anatomical
> image ? Thank you for your helptao----- Original Message ----- From:
> xuhaibo To: [log in to unmask] Sent: Wednesday, July 09, 2008 6:01
> AMSubject: Re: [SPM] realign_and_unwarp.mDear Julie,The "realign &
> unwarp " in the preprocessing is independent of brain shape or size,
> but normalization will be invloved in the brain size or shape due to
> rescaling. The "realign and unwarp" in pre-processinghas nothing to do
> with the normalization as mentioned by Karste n. It is better way, as
> my understanding, to use bias-corrected T1 from segmentation ( or
> *_seg_sn.mat file) as parameter reference for spatial normalization.
> Haibo> Date: Tue, 8 Jul 2 008 21:50:59 -0400> From: [log in to unmask]>
> Subject: Re: [SPM] realign_and_unwarp.m> To: [log in to unmask]> >
> Thank-you for your reply.> > The reason I inquired about realign &
> unwarp is that I am trying to> determine the optimal spatial
> preprocessing method for fMRI data acquired> from 2 groups: healthy
> controls and patients w/ Huntington's Disease.> T1-weighted images
> were also acquired.> > My concern is that the patients' brains may be
> structurally different from> controls. After realign & unwarp, what
> would be the best way to estimate the> warp parameters to use to write
> normalize the functional images into MNI> space? If I use the
> parameters produced by segmentatio n of the T1 image,> would the
> functionals be in MNI space after write normalization? If I use>
> normalize: estimate to determine the warp parameters and use the>
> bias-correcte d T1 from segmentation in addition to an MNI template,
> would I> use the mean functional image from realignment as the source
> image?> > I may be making this more complicated than it needs to be.
> Your advice would> be greatly appreciated.> > Regards,> > Julie> >
> -----Original Message-----> From: Dr. Karsten Mueller
> [mailto:[log in to unmask]] > Sent: Tuesday, July 08, 2008 12:06 PM>
> To: JULIE E MCENTEE> Subject: Re: [SPM] pixdim4 in NIfTI header after
> realign_anbd_unwarp.m> > Dear Julie,> > thank you for your email. In
> SPM5, you can decide if you want "realign> and reslice" or "realign
> and unwarp". If you decide for unwarp, SPM is> calculating the EPI
> distortions from the images. Alternatively, you can> specify a "voxel
> displacement map (vdm)" that describes the EPI> distortions. Thus, you
> can do the EPI distortion correctio n while> applying the realign
> parameters to the data set.> > Those things (realign and unwarp) are
> the first step in pre-processing> and have nothing to do with the
> normalization (into the MNI space). This> normalization can be done
> *after* the realignment.> > With best wishes to Baltimore,> Karsten> >
> JULIE E MCENTEE wrote:> > Dr. Mueller,> >> > I am interested in the
> SPM5 preprocessing option 'unwarp'. Could you> please tell me why you
> decided to use it? Did you normalize your images or> would that be
> redundant when you use unwarp?> >> > Thank-you,> >> > Julie> >> >
> Julie E. McEntee, M.A., C.C.R.P.> > Senior Research Program
> Coordinator> > Department of Psychiatry- Neuroimaging> > Johns Hopkins
> University School of Medicine> > 600 N. Wolfe St./ Phipps 300> >
> Baltimore, MD 21287> > Phone: 410-502-0468> > Fax: 410-6 14-3676> >> >
> ----- Original Message -----> > From: "Dr. Karsten Mueller" > > Date:
> Tuesday, July 8, 2008 5:08 am> > Subject: [SPM] pixdim4 in NIfTI
> header after realign_anbd_unwarp.m> > To: [log in to unmask]> >> >> >
> > >> Dear Colleagues,> >> > >> we are using Realign and Unwarp
> (spm_realign.m, spm_uw_estimate.m, and> >> spm_uw_apply.m) with
> 4D-NIfTI data which gives very nice results.> >> > >> However, these
> m-routines seem to change the NIfTI-header. Before> >> starting the
> realignment, our header sho ws> >> > >> time_units ms> >> pixdim4
> 2000.0000000000> >> > >> which is correct. After using realign and
> unwarp, the NIfTI header shows> >&g t; > >> time_units s> >> pixdim4
> 1.0000000000> >> > >> This might be a problem when further programs
> are using the repetition> >> time from the NIfTI header.> >> > >> With
> best wishes,> >> Karsten > >>
> ------------------------------------------------------------------------------ÇáËÉ°ÑHotmailÏÂÔص½±¾µØ£¬ÊÔÊÔ
> Windows Live Mail¡£ Á¢¼´³¢ÊÔ£¡ ===========End of original
> message text======================End of original message text===========
> _________________________________________________________________
> 这里好多好玩的视频,用 点到视频看看,有惊喜!
> http://cnweb.search.live.com/video/results.aspx?q=%E5%A5%A5%E8%BF%90%E5%9C%A3%E7%81%AB&mkt=zh-cn&FORM=WLMTWC===========End
> of original message text===========
|