Hi,
If you are just testing the mean activation, then you need to use the -ise
option to allow sign flipping. Otherwise, by default, PALM does permutations,
but permutations of your model will always generate the exact same mean, which
explains your results.
Cheers,
-MH
--
Michael Harms, Ph.D.
-----------------------------------------------------------
Associate Professor of Psychiatry
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave. Tel: 314-747-6173
St. Louis, MO 63110 Email: mha...@wustl.edu
From: <hcp-users-boun...@humanconnectome.org> on behalf of Gail Rosenbaum
<gailrosenb...@nyu.edu>
Date: Monday, July 2, 2018 at 4:42 PM
To: NEUROSCIENCE tim <tsc...@mst.edu>
Cc: hcp-users <hcp-users@humanconnectome.org>
Subject: Re: [HCP-Users] Blank p-value maps after running PALM on CIFTI files
Hi Tim,
Thanks for the clarification.
Here are the calls to palm for the separated subcortical and cortical files:
palm -i data_sub.nii -d ${StudyFolder}Level3/L3Setup.mat -t
${StudyFolder}Level3/L3Setup.con -o mbmfResults${contrast}_vol -T -logp
palm -i data_L.func.gii -d ${StudyFolder}Level3/L3Setup.mat -t
${StudyFolder}Level3/L3Setup.con -o mbmfResults${contrast}_SurfL -T -tfce2D -s
${StudyFolder}Level3/mbmfGroup_L.midthick.surf.gii
${StudyFolder}Level3/L_area.func.gii -logp
palm -i data_R.func.gii -d ${StudyFolder}Level3/L3Setup.mat -t
${StudyFolder}Level3/L3Setup.con -o mbmfResults${contrast}_SurfR -T -tfce2D -s
${StudyFolder}Level3/mbmfGroup_R.midthick.surf.gii
${StudyFolder}Level3/R_area.func.gii -logp
For this analysis, I am simply trying to take the mean of the activation to see
if any clusters are significantly greater than baseline. The design matrix
(L3Setup.mat) is a 36x1 matrix (there are 36 subjects in my current analysis)
and the t- L3Setup.con is just a 1. Maybe I set this up incorrectly?
Thanks,
Gail
On Mon, Jul 2, 2018 at 5:27 PM, Timothy Coalson
<tsc...@mst.edu<mailto:tsc...@mst.edu>> wrote:
As you have found, the outputs from PALM contain only 0s, therefore using a
different wb_command operation to convert them will not help with your current
issue.
If you can post the PALM call you used, it might help others who know PALM
figure out what your problem might be.
Tim
On Mon, Jul 2, 2018 at 4:12 PM, Gail Rosenbaum
<gailrosenb...@nyu.edu<mailto:gailrosenb...@nyu.edu>> wrote:
Hi Tim,
Thank you for your quick reply.
I renamed the maps and confirmed that there are no values in the maps (I pasted
the output of wb_command -file-information for the left hemisphere output below
in case that is useful).
Regarding -cifti-create-dense-from-template: I used this command in the first
iteration of my analyses and had the same issue. I switched to
-cifti-create-dense-scalar to see if it would work (and because this is the
suggested command and usage in the "HCP Practical 11: Task fMRI Analyses" pdf)
and again did not have any luck. That being said, I'm new to CIFTI files so I
appreciate the usage suggestions, and I will try -cifti-create-dense-scalar and
specify ROIs, but I am not sure this will help this issue because using a full
template produced the same issue.
Thanks,
Gail
Output of wb_command -file-information:
Name: mbmfResultsAllRPE3_SurfL_dpv_tstat_uncp.func.gii
Type: Metric
Structure: CortexLeft
Data Size: 129.97 Kilobytes
Maps to Surface: true
Maps to Volume: false
Maps with LabelTable: false
Maps with Palette: true
All Map Palettes Equal: true
Map Interval Units: NIFTI_UNITS_UNKNOWN
Number of Maps: 1
Number of Vertices: 32492
Map Minimum Maximum Mean Sample Dev % Positive % Negative
Inf/NaN Map Name
1 0.000 0.000 0.000 0.000 0.000 0.000
0 #1
On Mon, Jul 2, 2018 at 4:45 PM, Timothy Coalson
<tsc...@mst.edu<mailto:tsc...@mst.edu>> wrote:
Please name your tstat and p-value outputs ending in .func.gii, not simply .gii
. This will allow you to open them in wb_view directly, allowing you to do
sanity checks with fewer intervening steps. However, from what you have said,
I am fairly certain they are entirely 0s (you can use wb_command
-file-information to check the range of the maps, and any non-numeric values,
once you have named your outputs ending in .func.gii). I am not sure how to
check whether PALM was called with meaningful inputs that would be expected to
give significant results.
On a side note, by using -cifti-create-dense-scalar without specifying ROIs for
right and left cortex, you have generated CIFTI files that include the medial
wall, and do not match the 91k grayordinates of HCP data. I am guessing that
your subcortical labels are in fact the same as the HCP 91k grayordinate
subcortical labels, so I suggest you use -cifti-create-dense-from-template,
which makes it easier to match an existing specification of grayordinates.
Tim
On Mon, Jul 2, 2018 at 3:17 PM, Gail Rosenbaum
<gailrosenb...@nyu.edu<mailto:gailrosenb...@nyu.edu>> wrote:
Hi,
I am trying to run PALM with TFCE on dense timeseries CIFTI files created
through an HCP-style protocol, and preprocessed using the HCP pipeline (version
3.22). I am running these analyses on an HPC using SLURM.
It seems that I successfully ran PALM (there were no warnings or errors). When
I use -cifti-create-dense-scalar to merge the subcortical and cortical output
files, my t-statistic maps look fine in wb_view. However, when I use the same
code to merge the fwep and uncp maps and open them in wb_view, they appear to
be completely empty (and the range of values when I turn on the color bar is
just 0). I have tried using workbench versions 1.2.3 and 1.3.0.
For reference, here is the command that I used to successfully merge the t-stat
files:
wb_command -cifti-create-dense-scalar
mbmfGroup${contrast}_results_merged_tfce_tstat.dscalar.nii -volume
mbmfResults${contrast}_vol_tfce_tstat.nii data_sub_label.nii -left-metric
mbmfResults${contrast}_SurfL_tfce_tstat.gii -right-metric
mbmfResults${contrast}_SurfR_tfce_tstat.gii
And here is the command I used for the (seemingly) unsuccessful p-value maps:
wb_command -cifti-create-dense-scalar
mbmfGroup${contrast}_results_merged_tfce_tstat_fwep.dscalar.nii -volume
mbmfResults${contrast}_vol_tfce_tstat_fwep.nii data_sub_label.nii -left-metric
mbmfResults${contrast}_SurfL_tfce_tstat_fwep.gii -right-metric
mbmfResults${contrast}_SurfR_tfce_tstat_fwep.gii
My first thought was that there were no significant results, but since the
t-maps seem reasonable and even the uncorrected p-maps are empty (presumably if
nothing were signifiant, these uncorrected p-values would not 0), I think
something must have gone wrong.
Any insights into what might be going on would be appreciated! Thank you!
Gail Rosenbaum
--
Gail Rosenbaum, PhD
Postdoctoral Fellow | Hartley Lab
New York University
_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org>
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
--
Gail Rosenbaum, PhD
Postdoctoral Fellow | Hartley Lab
New York University
--
Gail Rosenbaum, PhD
Postdoctoral Fellow | Hartley Lab
New York University
_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
________________________________
The materials in this message are private and may contain Protected Healthcare
Information or other information of a sensitive nature. If you are not the
intended recipient, be advised that any unauthorized use, disclosure, copying
or the taking of any action in reliance on the contents of this information is
strictly prohibited. If you have received this email in error, please
immediately notify the sender via telephone or return mail.
_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
