Re: [HCP-Users] Table indicating location of clusters according to a dlabel file?

2017-09-07 Thread Timothy Coalson 
The commands in wb_command are designed for scripting flexibility, they
each do a small, low-level operation, to be chained together to achieve
various tasks.  However, they mainly output data files, there isn't much
for text output currently.

You could use -cifti-parcellate to parcellate your cluster maps, and any
parcel with a nonzero value therefore has some overlap - you can view this
file on the surface and click on any nonzero patch to check what area it
is.  You can also dump those values to text with -cifti-convert -to-text.
Running -file-information on the parcellated file or the dlabel file will
give you the order of the parcel names.

We aren't big fans of thresholding, and we would also consider parcellating
the timeseries before running the statistics, if your question is "which of
this parcellation's areas are significantly activated?".

Tim


On Thu, Sep 7, 2017 at 3:59 PM, Xavier Guell Paradis 
wrote:

> Dear HCP experts,
> I have a thresholded functional connectivity map (dscalar), and the dlabel
> files from the Glasser 2016 multimodal cortical parcellation. I was
> wondering whether there is a wb_command that would automatically generate a
> table indicating which labels overlap with my functional connectivity map.
> I have been exploring the wb_command index as well as the HCP mail archive
> and cannot find anything like this.
>
> Thank you very much,
> Xavier.
>
> ___
> HCP-Users mailing list
> HCP-Users@humanconnectome.org
> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

[HCP-Users] Table indicating location of clusters according to a dlabel file?

2017-09-07 Thread Xavier Guell Paradis 
Dear HCP experts,
I have a thresholded functional connectivity map (dscalar), and the dlabel 
files from the Glasser 2016 multimodal cortical parcellation. I was wondering 
whether there is a wb_command that would automatically generate a table 
indicating which labels overlap with my functional connectivity map.
I have been exploring the wb_command index as well as the HCP mail archive and 
cannot find anything like this.

Thank you very much,
Xavier.

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Re: [HCP-Users] Table indicating location of clusters according to a dlabel file?

2017-09-07 Thread Glasser, Matthew 
Right.  Basically we are suspicious of defining areas based on statistical 
thresholds, as these are unlikely to reflect biological boundaries in the 
brain, but rather the vagaries of the statistical thresholding approach and the 
noise distribution.

Peace,

Matt.

From: 
>
 on behalf of Timothy Coalson >
Date: Thursday, September 7, 2017 at 4:22 PM
To: Xavier Guell Paradis >
Cc: "hcp-users@humanconnectome.org" 
>
Subject: Re: [HCP-Users] Table indicating location of clusters according to a 
dlabel file?

The commands in wb_command are designed for scripting flexibility, they each do 
a small, low-level operation, to be chained together to achieve various tasks.  
However, they mainly output data files, there isn't much for text output 
currently.

You could use -cifti-parcellate to parcellate your cluster maps, and any parcel 
with a nonzero value therefore has some overlap - you can view this file on the 
surface and click on any nonzero patch to check what area it is.  You can also 
dump those values to text with -cifti-convert -to-text.  Running 
-file-information on the parcellated file or the dlabel file will give you the 
order of the parcel names.

We aren't big fans of thresholding, and we would also consider parcellating the 
timeseries before running the statistics, if your question is "which of this 
parcellation's areas are significantly activated?".

Tim


On Thu, Sep 7, 2017 at 3:59 PM, Xavier Guell Paradis 
> wrote:
Dear HCP experts,
I have a thresholded functional connectivity map (dscalar), and the dlabel 
files from the Glasser 2016 multimodal cortical parcellation. I was wondering 
whether there is a wb_command that would automatically generate a table 
indicating which labels overlap with my functional connectivity map.
I have been exploring the wb_command index as well as the HCP mail archive and 
cannot find anything like this.

Thank you very much,
Xavier.

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users


___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Re: [HCP-Users] Table indicating location of clusters according to a dlabel file?

2017-09-07 Thread Xavier Guell Paradis 
Thank you Matt and Tim for the very useful comments!
Xavier.

From: Glasser, Matthew [glass...@wustl.edu]
Sent: Thursday, September 07, 2017 5:25 PM
To: NEUROSCIENCE tim; Xavier Guell Paradis
Cc: hcp-users@humanconnectome.org
Subject: Re: [HCP-Users] Table indicating location of clusters according to a 
dlabel file?

Right.  Basically we are suspicious of defining areas based on statistical 
thresholds, as these are unlikely to reflect biological boundaries in the 
brain, but rather the vagaries of the statistical thresholding approach and the 
noise distribution.

Peace,

Matt.

From: 
>
 on behalf of Timothy Coalson >
Date: Thursday, September 7, 2017 at 4:22 PM
To: Xavier Guell Paradis >
Cc: "hcp-users@humanconnectome.org" 
>
Subject: Re: [HCP-Users] Table indicating location of clusters according to a 
dlabel file?

The commands in wb_command are designed for scripting flexibility, they each do 
a small, low-level operation, to be chained together to achieve various tasks.  
However, they mainly output data files, there isn't much for text output 
currently.

You could use -cifti-parcellate to parcellate your cluster maps, and any parcel 
with a nonzero value therefore has some overlap - you can view this file on the 
surface and click on any nonzero patch to check what area it is.  You can also 
dump those values to text with -cifti-convert -to-text.  Running 
-file-information on the parcellated file or the dlabel file will give you the 
order of the parcel names.

We aren't big fans of thresholding, and we would also consider parcellating the 
timeseries before running the statistics, if your question is "which of this 
parcellation's areas are significantly activated?".

Tim


On Thu, Sep 7, 2017 at 3:59 PM, Xavier Guell Paradis 
> wrote:
Dear HCP experts,
I have a thresholded functional connectivity map (dscalar), and the dlabel 
files from the Glasser 2016 multimodal cortical parcellation. I was wondering 
whether there is a wb_command that would automatically generate a table 
indicating which labels overlap with my functional connectivity map.
I have been exploring the wb_command index as well as the HCP mail archive and 
cannot find anything like this.

Thank you very much,
Xavier.

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users


___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

[HCP-Users] rfMRI processing

2017-09-07 Thread hercp 
I am looking at rfMRI data.  Are there files that are already time-filtered 
between .01 and .08 or should I do the filtering myself?  If there are, where 
are they, and how are they labeled?

Thanks once again,
Heracles Panagiotides, PhD



___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Re: [HCP-Users] rfMRI processing

2017-09-07 Thread Timothy Coalson 
We do not apply that kind of narrow temporal filtering (we basically only
do a detrend), you will need to do such things yourself.

Tim


On Thu, Sep 7, 2017 at 9:27 PM, hercp  wrote:

> I am looking at rfMRI data.  Are there files that are already
> time-filtered between .01 and .08 or should I do the filtering myself?  If
> there are, where are they, and how are they labeled?
>
> Thanks once again,
> Heracles Panagiotides, PhD
>
>
> ___
> HCP-Users mailing list
> HCP-Users@humanconnectome.org
> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Re: [HCP-Users] rfMRI processing

2017-09-07 Thread Glasser, Matthew 
Also, I wouldn’t recommend doing that.  There is plenty of BOLD signal outside 
those frequencies.

Peace,

Matt.

From: 
>
 on behalf of Timothy Coalson >
Date: Thursday, September 7, 2017 at 9:33 PM
To: hercp >
Cc: HUMAN CONNECTOME 
>
Subject: Re: [HCP-Users] rfMRI processing

We do not apply that kind of narrow temporal filtering (we basically only do a 
detrend), you will need to do such things yourself.

Tim


On Thu, Sep 7, 2017 at 9:27 PM, hercp > wrote:
I am looking at rfMRI data.  Are there files that are already time-filtered 
between .01 and .08 or should I do the filtering myself?  If there are, where 
are they, and how are they labeled?

Thanks once again,
Heracles Panagiotides, PhD



___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users


___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users