[caret-users] listing clusters

[donna at brainvis.wustl.edu]

Just another remark:  When you run a Surface: ROI report as DVE describes
below, you will get not ony the ROI center of gravity coordinates (in this
case, entire cortical surface), but also a list of mean, stddev, min, and
max for each metric and/or surface shape column you have loaded.  From
this list, you can get the max for your entire ROI.  Then, you can use the
same Surface: ROI feature to generate a surface shape cluster report. 
You'll need to enter that max value in the threshold box.  You'll probably
need to load your metric as a surface_shape file for this to work.

This will work only for the max of the maxes -- you won't get local maxima
this way.  But if you click around on your metric, the Surface: ROI
feature also has a feature where you can limit the selection to nodes of a
given threshold connected to a node you click on.

> Christa-Lynn,
>
> Two comments.
>
> 1) Working from the menu options, you can use the Surface: Region of
> Interest window to do something that I think is close to what you want.
>
> Select your overall ROI.  If, for example, you want everything except
> the medial wall, choose
> Selection Method: Nodes with Paint;
> Category: Medial Wall;
> Select Name: Medial Wall;
> Select Nodes
> Invert Node Selection.
>
> Then, with the metric file copied to have a *.surface shape file
> ending, load it as a surface shape file and view the relevant column
> in the main window surface.
>
> Select Surface ROI: Operate on Selected Nodes: Operation: Surface
> Shape Cluster Analysis, and set the threshold where desired.
>
> The Report column will show the surface area and the center of gravity
> of each cluster.
>
> You said that you really want the maximum value of each cluster and
> the coordinates of that maximum.  I'm not aware of a Caret function
> that does that. However, the request is a reasonable one, and if it is
> indeed what you want we can add it to John Harwell's to-do list
> (presumably better to operate on metric files directly).  One
> important issue would be what to do about ties - it will not be
> uncommon for multiple nodes within a cluster to have the 'maximum'
> value.  If so, should all of them be reported?  (could be a long
> list)  Or should the COG of all 'peak nodes' within a cluster be
> reported?  Or some other sensible strategy?
>
> It would help to get feedback on this issue and also whether other
> Caret users would find this to be a useful analysis function.
>
> 2) A separate matter is that whenever I encounter someone still using
> our 'old' Colin atlas, I take the opportunity to recommend strongly
> that you switch to the new PALS human atlas, which we consider to be
> superior on many grounds.  For the background and rationale, see:
>
> Van Essen & Dierker, Surface-based and probabilistic atlases of
> primate cerebral cortex. Neuron. 2007;56:209-25
> http://www.ncbi.nlm.nih.gov/pubmed/17964241
>
> and/or:
> Van Essen A Population-Average, Landmark- and Surface-based (PALS)
> atlas of human cerebral cortex.
> Neuroimage. 2005  28:635-62
> http://www.ncbi.nlm.nih.gov/pubmed/16226584
>
> For the atlas dataset and tutorial, see
> http://sumsdb.wustl.edu/sums/directory.do?id=6585200&dir_name=CARET_TUTORIAL_SEPT-06
>
> Good luck!
>
> David VE
>
>
> On Jan 10, 2008, at 6:27 PM, donovan at psych.ucsb.edu wrote:
>
>> Hi,
>> I have a question on how to get a list of clusters from a
>> thresholded metric file. I am displaying SPM2 t images (fMRI) onto
>> the human.colin.Cerebral atlas in Caret v5.3. I am able to threshold
>> the image but I can't seem to figure out how to get a list of maxima
>> and their coordinates. I found in your archives a suggestion to copy
>> the metric file to a surface_shape but I am unfamiliar with the
>> command line prompts you suggested (I have only used the menus).
>> Any help you can give me with doing this through the menus would be
>> greatly appreciated.
>>
>> Thanks in advance!
>>
>> Christa-Lynn Donovan
>> PhD Candidate
>> Department of Psychology
>> University of California
>> Santa Barbara, 93106
>>
>>
>> _______________________________________________
>> caret-users mailing list
>> caret-users at brainvis.wustl.edu
>> http://pulvinar.wustl.edu/mailman/listinfo/caret-users
>
> _______________________________________________
> caret-users mailing list
> caret-users at brainvis.wustl.edu
> http://pulvinar.wustl.edu/mailman/listinfo/caret-users
>