That was a slip of the keyboard (I sent 24 emails to the list that day). I
meant to say "You would need to segment the T1 image".
Best regards,
-John
> I am having difficulties to spatially normalize IR MR images to T1.mnc in
> SPM2. I am spatially normalizing PET data using coregistered MRI. The
> source of the problem might be that my MR data have very high contrast
> between gray and white matter and a little noisy. By using my IR, T2 and
> Flair images together, segmentation works well. Using the segmented gray
> matter images, spatial normalization to gray.mnc works well. But I wanted
> to know if there was a problem to do so. By searching the SPM Mailbase, I
> found the same question as copied below. I am confused with the answer by
> Dr. John Ashburner. Did he suggest segmenting T1.mnc to create a template
> image of gray matter? It is unlikely to work because T1.mnc is smoothed.
> Wasn't gray.mnc created from the T1 template before smoothing? Is there a
> problem to normalize my segmented gray matter images to gray.mnc?
>
> Date: Mon, 29 Mar 2004 16:25:24 +0000
>
> Reply-To: John Ashburner <[log in to unmask]>
>
> Sender: "SPM (Statistical Parametric Mapping)" <[log in to unmask]>
>
> From: John Ashburner <[log in to unmask]>
>
> Organization: FIL
>
> Subject: Fwd: Re: [SPM] normalization with gray.mnc
>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Content-Disposition: inline
>
>
>
> ---------- Forwarded Message ----------
>
>
>
> Subject: Re: [SPM] normalization with gray.mnc
>
> Date: Thursday 25 March 2004 11:35
>
> From: John Ashburner <>
>
> To: Nici Wenderoth <[log in to unmask]>, [log in to unmask]
>
> > we had some problems with normalizing 3T EPIs, because we often observed
> >
> > activation just outside the glasbrain. We've tried to segment our
> >
> > anatomical image and using the gray matter only. As a template image,
>
> we've
>
> > used the gray.mnc in the apriori directory. Is this a good choice, I mean
> >
> > can we really use the gray.mnc or do we have to segment the T1 template
> >
> > image???
>
> You would need to segment the T1 template if you plan to match grey matter
>
> with the gray.mnc image. Spatial normalisation in SPM2 minimises the mean
>
> squared difference between the image and template, so needs a template with
>
> similar contrasts between the different tissue types.
>
>
>
> Best regards,
>
> -John
|