Hi Mark,
yes - that clarifies it. Thank you!
Cheers,
Niklas
PS. I was confused - but in an optimistic way, therefore 'amazed'
Mark Jenkinson wrote:
> Hi,
>
> The command cluster does both - it calculates the cluster
> images and it can also perform GRF calculations. I'm not
> sure why this is amazing. If you are going to do GRF
> calculations then you need to give is a smoothness
> estimate via the option:
> -d,--dlh smoothness estimate = sqrt(det(Lambda))
> which you can get from the program smoothest.
>
> In order to get the GRF output you need to produce an
> appropriate output such as --opvals or --minclustersize
>
> I hope this is clearer.
> All the best,
> Mark
>
>
> On 17 Feb 2010, at 08:57, Niklas Stein wrote:
>
>> Hi Steve,
>>
>> thanks for your reply. My question refers to the command
>> '$FSLDIR/bin/cluster'. My impression is that it just puts out a cluster
>> list of (face, edge or vertex?) adjacent voxels, but the website states
>> that it uses GRF, which I am amazed of. Therefore, I was wondering
>> whether the output from e.g. 'cluster -i stat_vox_p_tstat<n> -t .95'
>> includes any kind of GRF based correction or if the website is
>> misleading here.
>>
>> Cheers,
>> Niklas
>>
>> Stephen Smith wrote:
>>> Hi - there's a little more info in the FEAT manual
>>> http://fsl.fmrib.ox.ac.uk/fsl/feat5/detail.html#poststats
>>> FEAT offers both voxel-based and cluster-based thresholding, in both
>>> cases the test statistic (voxel Z value and cluster size
>>> respectively) are turned into corrected p-values using GRF theory.
>>> Not sure if that answers your question - but hopefully it's a pointer
>>> in the right direction. For a lot more useful stuff on GRF see the
>>> Cambridge Imagers website.
>>> Cheers.
>>> On 16 Feb 2010, at 12:02, Niklas Stein wrote:
>>>> Hi experts,
>>>>
>>>> on the individual tools' documentation pages, for inference
>>>> 'cluster' is
>>>> documented as 'cluster-based thresholding using GRF theory for
>>>> inference'. I was wondering, how (and why) GRF is incorporated into
>>>> 'cluster' and whether 'cluster' has any other purpose than giving a
>>>> list
>>>> of clusters by thresholding output from randomise.
>>>> Please do not laugh at the potential sillyness of my question, but
>>>> since
>>>> the documentation is sparse and it is listed as an inference tool I
>>>> feel
>>>> in need for clarification.
>>>>
>>>> Cheers,
>>>> Niklas
>>>>
>>>>
>>>> --
>>>> Pflichtangaben gemäß Gesetz über elektronische Handelsregister und
>>>> Genossenschaftsregister sowie das Unternehmensregister (EHUG):
>>>>
>>>> Universitätsklinikum Hamburg-Eppendorf
>>>> Körperschaft des öffentlichen Rechts
>>>> Gerichtsstand: Hamburg
>>>>
>>>> Vorstandsmitglieder:
>>>> Prof. Dr. Jörg F. Debatin (Vorsitzender)
>>>> Dr. Alexander Kirstein
>>>> Prof. Dr. Dr. Uwe Koch-Gromus
>>> ---------------------------------------------------------------------------
>>>
>>> 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
>>> ---------------------------------------------------------------------------
>>>
>>
>>
>> --
>> ----------------------------------------------------------------------
>> Dipl.-Phys. Niklas Stein
>> ----------------------------------------------------------------------
>> Institut für Systemische Neurowissenschaften
>> Universitätsklinikum Hamburg-Eppendorf
>> Martinistr. 52, Haus W34
>> D-20246 Hamburg, Germany
>> Tel.: +49-40-7410-57882
>> Fax.: +49-40-7410-59955
>> mail: [log in to unmask]
>> http:/www.uke.uni-hamburg.de/institute/systemische-neurowissenschaften
>> ----------------------------------------------------------------------
>> Time ferðast mun hafa gert flóknari málfræði.
>> ----------------------------------------------------------------------
>> () ascii ribbon campaign (http://arc.pasp.de)
>> /\ * against html mail and vcards *
>> ----------------------------------------------------------------------
>> Please avoid sending me Microsoft Office files as attachments.
>> (see http://www.gnu.org/philosophy/no-word-attachments.html)
>> ----------------------------------------------------------------------
--
Pflichtangaben gemäß Gesetz über elektronische Handelsregister und Genossenschaftsregister sowie das Unternehmensregister (EHUG):
Universitätsklinikum Hamburg-Eppendorf
Körperschaft des öffentlichen Rechts
Gerichtsstand: Hamburg
Vorstandsmitglieder:
Prof. Dr. Jörg F. Debatin (Vorsitzender)
Dr. Alexander Kirstein
Prof. Dr. Dr. Uwe Koch-Gromus
|