Hi - our general experience is that two-channel segmentation can give
improved subcortical segmentation, but that T1-only segmentation can
give better cortical segmentation. We've been meaning to knock up a
script to merge the two approaches using standard-space masking and
merging, but now that we have FIRST, there's less reason to put
effort into worrying about subcortical segmentation with FAST.
Cheers.
On 8 Sep 2007, at 19:50, Jamie Hanson wrote:
>> Hi - the FAST segmentation includes bias field correction within it,
>> so you don't need to worry about this.
>>
>> The intensity range scaling is unrelated to bias field. The
>> segmentation in FAST is not sensitive to the absolute scaling, so you
>> can probably get away with just working out the scaling for the
>> greatest-range image and using that for all inputs, as long as the
>> smallest intensity images are not a _tiny_ fraction of the greatest.
>
> I knew FAST had a bias correction algorithm in it, but i was under the
> (possibly false?) impression that doing a 2-channel bias correction
> with MFAST, would yield a better bias correction and possibly better
> segments (than the 1 channel FAST)?
>
> I have T2s and T1s, so in the end, would it yield a substantially
> better study-specific template, segments, VBM-results, etc. etc., to
> do an initial 2-channel MFAST bias correction, then put those
> corrected T1s into the various FSL-VBM pathways?
>
>
> thanks much,
> jamie.
>
>
>
>
> On 9/8/07, Steve Smith <[log in to unmask]> wrote:
>> Hi - the FAST segmentation includes bias field correction within it,
>> so you don't need to worry about this.
>>
>> The intensity range scaling is unrelated to bias field. The
>> segmentation in FAST is not sensitive to the absolute scaling, so you
>> can probably get away with just working out the scaling for the
>> greatest-range image and using that for all inputs, as long as the
>> smallest intensity images are not a _tiny_ fraction of the greatest.
>>
>> Cheers.
>>
>>
>>
>>
>> On 7 Sep 2007, at 18:24, Jamie Hanson wrote:
>>
>>> Hello all-
>>>
>>> I am a newbie to the list, so I apologize if this question has been
>>> recently answered (I searched the archives, but didn't see any
>>> answers
>>> right off the bat).
>>>
>>> But I was wondering if you are using the basic analytic pathway
>>> specified in the "FSL-VBM" approach, whether you should bias correct
>>> first (before creating a study specific template, etc)?
>>>
>>> Related to this, I have been running fslstats and finding the
>>> need to
>>> use greatly different scaling parameters. For some T1s, the
>>> ranges are
>>> from 0 to 6000, while other are 0 13000 (and every value in
>>> between).
>>> There is definitely no consistent scaling factor I could apply. Does
>>> that mean bias correction is needed first?
>>>
>>> Thoughts? Suggestions?
>>>
>>> Best,
>>> jamie.
>>>
>>>
>>>
>>>
>>> --
>>> Jamie L. Hanson
>>> Waisman Laboratory for Brain Imaging & Behavior | Child Emotion
>>> Research Lab
>>> University of Wisconsin - Madison
>>> 1500 Highland Avenue
>>> Madison, WI 53706
>>> Phone: (608) 262-5148
>>> ***Please note my new email: [log in to unmask]***
>>
>>
>> ---------------------------------------------------------------------
>> ---
>> ---
>> Stephen M. Smith, Professor of Biomedical Engineering
>> Associate Director, Oxford University FMRIB Centre
>>
>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
>> +44 (0) 1865 222726 (fax 222717)
>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
>> ---------------------------------------------------------------------
>> ---
>> ---
>>
>
>
> --
> Jamie L. Hanson
> Waisman Laboratory for Brain Imaging & Behavior | Child Emotion
> Research Lab
> University of Wisconsin - Madison
> 1500 Highland Avenue
> Madison, WI 53706
> Phone: (608) 262-5148
> ***Please note my new email: [log in to unmask]***
------------------------------------------------------------------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
------------------------------------------------------------------------
---
|