open-discussion
open-discussion > RE: Problem with SPHARM_Tutorial_Data_July2015
Jun 14, 2016 10:06 PM | Beatriz Paniagua
RE: Problem with SPHARM_Tutorial_Data_July2015
Hi Shahrokh,
Sorry for the delay in responding this question.
Is there a relationship between signed distance map and p-value map (e.g., mancovaRawP)?
Not necessarily. The p-value maps illustrate significantly
different local shape changes between the two populations
introduced in mancova. This is not related to the magnitude of the
measurement.
I expected that there would be a direct relationship between signed distance and p-value color-coded maps. In other words, is it expected that p-values in the specific region, relation with other regions, reduces as signed magnitudes of the distance in this specific region (between mean surface of normal group and mean surface of OA group) is larger than other regions?
Again, not necessarily. See response above.
I hope that helps.
Please, let me know if you have any further questions.
Best,
Beatriz
Sorry for the delay in responding this question.
Is there a relationship between signed distance map and p-value map (e.g., mancovaRawP)?
I expected that there would be a direct relationship between signed distance and p-value color-coded maps. In other words, is it expected that p-values in the specific region, relation with other regions, reduces as signed magnitudes of the distance in this specific region (between mean surface of normal group and mean surface of OA group) is larger than other regions?
Can you analyze these images? Also, can you make
relevant these images together in terms of color?
Thanks for sharing your images. Large magnitudes of change between
the two means do not have to result in more significant changes. A
lot of things are in play for example variability of the
measurement in the region.The outputs of shapeAnalysisMANCOVA_Wizard are
meta files with the extensions as *_meanA.meta and *_meanB.meta.
How can created these
surfaces? Is this mean surface can be representative of its group?
These means are the results of averaging the coordinate points of
each one of the populations. The mean is just a way to describe
your group, but of course it mean a data reduction. If your group
includes a lot of shape variability it will be all averaged in the
end, and it wont be able to represent the group well. Just curious,
what is the n of each one of your groups?surfaces? Is this mean surface can be representative of its group?
The contents of files as
*_mancovaBonferroniP.txt or *_mancovaFDRP.txt is 1 exactly!
That means that your p-values do not survive correction for false
positives in the threshold you specified in the
shapenalysismancova_wizard program.I hope that helps.
Please, let me know if you have any further questions.
Best,
Beatriz
Threaded View
Title | Author | Date |
---|---|---|
shahrokh nasseri | Apr 4, 2016 | |
Beatriz Paniagua | Apr 10, 2016 | |
shahrokh nasseri | Apr 27, 2016 | |
shahrokh nasseri | May 10, 2016 | |
Beatriz Paniagua | May 10, 2016 | |
shahrokh nasseri | May 11, 2016 | |
Beatriz Paniagua | Apr 27, 2016 | |
shahrokh nasseri | May 1, 2016 | |
Beatriz Paniagua | May 2, 2016 | |
shahrokh nasseri | May 10, 2016 | |
shahrokh nasseri | May 16, 2016 | |
Beatriz Paniagua | May 16, 2016 | |
shahrokh nasseri | Jun 8, 2016 | |
Beatriz Paniagua | Jun 14, 2016 | |
shahrokh nasseri | Jun 16, 2016 | |