dHCP statistics

300 views
Skip to first unread message

Hobert, Leah E.

unread,
May 24, 2021, 11:24:30 AM5/24/21
to hcp-...@humanconnectome.org
Hello,

I would like to get statistics from a .nii.gz file for dHCP labels. What can I use to get statistics, such as thickness, volume, curvature, etc in connectome workbench?

Thanks,
Leah
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Mass General Brigham Compliance HelpLine at http://www.massgeneralbrigham.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.

Please note that this e-mail is not secure (encrypted).  If you do not wish to continue communication over unencrypted e-mail, please notify the sender of this message immediately.  Continuing to send or respond to e-mail after receiving this message means you understand and accept this risk and wish to continue to communicate over unencrypted e-mail. 

Coalson, Timothy Scott (S&T-Student)

unread,
May 24, 2021, 2:26:45 PM5/24/21
to hcp-...@humanconnectome.org
We use surface files (.surf.gii) to get thickness and curvature information, which for our data come from running the structural pipelines (the ones that have "freesurfer" in their name) on the T1w and T2w scans.  You can get per-vertex volume measurement with wb_command -surface-wedge-volume .

If you have these measures in cifti format (and the area labels you want to use as a matching cifti dlabel file), you can use -cifti-parcellate to average or sum the measures within areas.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Monday, May 24, 2021 10:10 AM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/DM5PR04MB0810C9A36B1257F717AD02ADD9269%40DM5PR04MB0810.namprd04.prod.outlook.com.

Harms, Michael

unread,
May 24, 2021, 2:58:36 PM5/24/21
to hcp-...@humanconnectome.org

 

We also have all the FS statistic outputs in <subjID>/T1w/<subjID>/stats

 

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: "Coalson, Timothy Scott (S&T-Student)" <tsc...@mst.edu>

Subject: [hcp-users] Re: dHCP statistics

 

* External Email - Caution *

 

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.

Hobert, Leah E.

unread,
May 28, 2021, 1:44:09 PM5/28/21
to hcp-...@humanconnectome.org
Hello,

I am following up on the question in my previous email. Is there a wb_command that can be used to get statistics from cifti or gifti files?

Also, for cifti-stats, what are the values used for -reduce and -percentile? What stats should be the output?

Thanks,
Leah
From: Hobert, Leah E.
Sent: Monday, May 24, 2021 11:10 AM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: dHCP statistics
 

Hobert, Leah E.

unread,
May 28, 2021, 1:50:52 PM5/28/21
to hcp-...@humanconnectome.org
Hello,

Thank you for your previous response. I did not see it earlier. Please disregard the first question.

Thanks,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, May 28, 2021 1:44 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: Re: dHCP statistics
 

Hobert, Leah E.

unread,
May 28, 2021, 3:38:37 PM5/28/21
to hcp-...@humanconnectome.org
Are there examples or more information on how to use wb_command? I am looking to get stats on only temporal labels and would like help understanding the parameters. For wb_command -surface-wedge-volume, should the parameters (inner-surface, outer-surface, metric) be file names or numbers? For wb_command-cifti-parcellate, what is the direction?

Also, I did not find the T1w directory. Would this be under the derivatives/sub-<subid>_ses-<sesid>/anat directory?
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, May 28, 2021 1:50 PM

Coalson, Timothy Scott (S&T-Student)

unread,
May 28, 2021, 4:24:52 PM5/28/21
to hcp-...@humanconnectome.org
-surface-wedge-volume expects filenames.  In general, anything in wb_command that uses files as input or output requires them as filenames.

vertex areas or vertex volumes converted to cifti should use the dscalar type.  As -cifti-parcellate mentions, for dscalar you should use "COLUMN".

wb_command contains some built-in help, particularly the -arguments-help option, which is also available as web pages:


There is also the section of the HCP course that covers some basics:



The path you mention looks like a BIDS structure.  I believe the dHCP project wrote their own pipelines, and did not use the HCP pipelines, so their outputs are not organized the same way (and may not have equivalents to some HCP pipeline outputs).

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, May 28, 2021 2:38 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 1, 2021, 3:04:59 PM6/1/21
to hcp-...@humanconnectome.org
Hi Tim,

Thanks for the references! I created a file with only the temporal labels from drawem.dlabel.nii and visualized the labels over the white and inflated surfaces. I am trying to get stats (thickness, volume, curvature) under the temporal labels. When I tried -surface-wedge-volume, it used inner and outer surf.gii files and the output was func.gii. For cifti-parcellate, the cifti-in and label files were .nii and the output was pscalar.nii. The functions created output files but I could not open them. What should the outputs be for -surface-wedge-volume and cifti-parcellate? How would I get stats from these files?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, May 28, 2021 3:38 PM

Hobert, Leah E.

unread,
Jun 1, 2021, 3:10:19 PM6/1/21
to hcp-...@humanconnectome.org
Also, would these functions or something else be more useful for getting a stats file from the label surface volume?
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Tuesday, June 1, 2021 3:04 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 1, 2021, 4:24:58 PM6/1/21
to hcp-...@humanconnectome.org
The wb_command -gifti-help and -cifti-help explanations may help here.  -surface-wedge-volume outputs a "metric" file.  You can use -cifti-create-dense-from-template to make a cifti file from metric files and a template cifti file (any 91k cifti file can be used as the template).  The cifti format has a way to represent parcellated data directly (and the gifti format does not), which is why there is a parcellate command for cifti (doing so on gifti would need to output a text file).

If you aren't interested in the answers for more than a handful of labels, and are fine with keeping track of which number is from which parcel manually, you can use -metric-stats or -metric-weighted-stats (vertex area or volume should be simply summed, thickness should be a weighted average by vertex area) multiple times with ROIs (single-column metric files with binary values, you can do -cifti-separate to get a .label.gii from the dlabel file, then -gifti-all-labels-to-rois and -metric-merge to split out the ones you need).

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Tuesday, June 1, 2021 2:10 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 3, 2021, 5:06:27 PM6/3/21
to hcp-...@humanconnectome.org
Hi Tim,

Could you clarify what the inner and outer surfaces are for -surface-wedge-volume? Would this have to be done separately for each hemisphere? Should I be able to view the output using wb_view? I used the following command to create the output and workbench crashed when I tried to open it.
wb_command -surface-wedge-volume sub-CC00117XX10_ses-38200_left_white.surf.gii sub-CC00117XX10_ses-38200_left_pial.surf.gii sub-CC00117XX10_ses-38200_vol.func.gii. 

For metric-stats, how would I get the value used for -reduce and -percentile? What stats should be the output?

I tried using metric-weighted-stats with the following command but got an error.
wb_command -metric-weighted-stats sub-CC00117XX10_ses-38200_left_drawem.label.gii -sum -area-surface sub-CC00117XX10_ses-38200_left_white.surf.gii
WARNING: Metric File: sub-CC00117XX10_ses-38200_left_drawem.label.gii contains data array with NIFTI_INTENT_LABEL !!!

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Tuesday, June 1, 2021 3:10 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 3, 2021, 5:52:16 PM6/3/21
to hcp-...@humanconnectome.org
Yes, it needs to be done separately per hemisphere.  You should use the pial and white surfaces, and the result should display in wb_view (you will also need to load surfaces to display it on).

Please reproduce the wb_view crash by running "wb_view" in a terminal and loading the file(s) into that new wb_view window, and tell us what error message you get (copy the text into an email, every line after the last "INFO" line).  Note that if you have surfaces loaded with a different resolution than the metric file(s) you tried to load, you will get an error message, but the wb_view window should not close (crash) because of it.

-metric-stats and -metric-weighted-stats are intended to be run on metric files, they do not understand label files, the output of that -metric-weighted-stats command will be nonsense (that message is a warning, not an error, because it is occasionally useful to treat the values in a label file as if they were a metric file, but in many cases it indicates you did something wrong).  Their output is a number that they output directly to the terminal.  The command -gifti-label-to-roi may be a simpler way to get just a selected handful of labels as ROI metric files.  Again, if you want to get values for all (or any large number) of areas, I would strongly suggest using cifti format and -cifti-parcellate.

I don't think -percentile is applicable for what you want to do, so don't specify this option.  The options you can use in -reduce are specified in the help info of the command.  For your purposes, you should use -reduce SUM for vertex areas or vertex volumes, while thickness should use -metric-weighted-stats with -mean, using the vertex areas as weights.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Thursday, June 3, 2021 4:06 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 4, 2021, 2:03:53 PM6/4/21
to hcp-...@humanconnectome.org
Hi Tim,

Please find the error below I got when wb_view crashed.

Info: Resources loaded:
   :/About   :/BalsaUploadDialog   :/Cursor   :/Fonts   :/general_resources.qrc   :/help_resources.qrc   :/HelpFiles   :/LayersPanel   :/PaletteSettings   :/qt-project.org   :/SceneFileDialog   :/SpecFileDialog   :/Splash   :/ToolBar   :/update_resources.sh

libGL error: unable to load driver: swrast_dri.so
libGL error: failed to load driver: swrast
/usr/pubsw/bin/wb_view: line 17: 28439 Segmentation fault      (core dumped) "$directory"/../exe_rh_linux64/wb_view "$@"

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Thursday, June 3, 2021 5:06 PM

Glasser, Matthew

unread,
Jun 4, 2021, 2:10:30 PM6/4/21
to hcp-...@humanconnectome.org

Are you able to use Connectome Workbench wb_view at all?

Matt.

--

You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 4, 2021, 2:25:50 PM6/4/21
to hcp-...@humanconnectome.org
Hi Matt,

Yes, I can use wb_view to open files in the DHCP directories. It just didn't work for the wb_command -surface-wedge-volume output.

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 4, 2021 2:03 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 4, 2021, 3:05:32 PM6/4/21
to hcp-...@humanconnectome.org
Did you launch wb_view differently when viewing the dHCP files?  From matlab or in a container?  That error is what I would expect when it can't find good 3D libraries, which shouldn't depend on what data files you try to load.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 4, 2021 1:25 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 4, 2021, 5:08:18 PM6/4/21
to hcp-...@humanconnectome.org
Hi Tim,

I launched wb_view the same way in a container. It worked when I opened a surface file first. I was able create a pscalar.nii file using cifti-parcellate and view it in workbench. How would I get stats for the labels from this?

If I used metric-stats, would I need to convert each label in the temporal ROI separately into a metric file or would it output a number for each label? Could I use cifti-stats instead on the label file?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 4, 2021 2:25 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 4, 2021, 5:20:53 PM6/4/21
to hcp-...@humanconnectome.org
You can either read the pscalar file into whatever language you want, or you can convert it to a text file of numbers with -cifti-convert.  You can get the order of the parcels from -cifti-label-export-table on the dlabel file you used, or -file-information on the pscalar file.

The -*-stats commands do not understand label files, so yes, the labels need to be converted to ROIs in order to use them in those commands.  However, it looks like the ROIs can be multiple maps in one file, and it will give you a line of numbers, one for each ROI.  I should update the help info on -metric-stats and -volume-stats...

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 4, 2021 4:08 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 7, 2021, 12:06:13 PM6/7/21
to hcp-...@humanconnectome.org
Hi Tim,

Thanks for the info. When the pscalar file is converted using cifti-convert, what do the numbers in the text file mean? I got a list of 10 numbers alternating with zeroes. I am looking to get information on the temporal labels, such as thickness, volume, and curvature.

When I used -file-information, I got a list of parcels and number of vertices. Is there a way to use the vertices to get the stats I listed above? Also, I made a dscalar file for the left cortex with -cifti-create-dense-from-template, which I used in -cifti-parcellate to make the pscalar file. Why does -file-information have vertices for both the left and right cortex?

For -cifti-label-export-table, how is the label map different from a label file?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 4, 2021 5:08 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 7, 2021, 5:02:25 PM6/7/21
to hcp-...@humanconnectome.org
-cifti-create-dense-from-template will have the same indices as whatever template file you used, so that is likely why you have both hemispheres.  Single-hemisphere cifti files (that don't have the usual 91k indices) would complicate comparison with other cifti data, so we don't typically recommend it.

The output of -cifti-convert is the data values in the cifti file, as-is.  I would guess that your label file uses labels "L_<something> 1, R_<something> 2", etc, resulting in an alternating pattern if you parcellated a file that contains zeros for the entire right hemisphere.

"<map>" refers to an index (or map name) in the file, you usually want to use "1" - dlabel files can store more than one parcellation per file.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Monday, June 7, 2021 11:06 AM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 9, 2021, 4:27:52 PM6/9/21
to hcp-...@humanconnectome.org
Hi Tim,

I still do not understand what the units are for the data values in -cifti-convert and -cifti-label-export-table. Are they the number of vertices?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Monday, June 7, 2021 12:06 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 9, 2021, 6:38:39 PM6/9/21
to hcp-...@humanconnectome.org
If the input to -cifti-parcellate was thickness in mm and you used the default reduction (mean), the units in the parcellated file, and therefore the output of -cifti-convert, are average thickness of each parcel, in mm.  If you used volume in mm^3 as the input to -cifti-parcellate, and used -reduce SUM, the units are volume of each parcel in mm^3.

-cifti-convert doesn't do anything special, it just takes the data values from one file and puts them as-is into a new file.

-cifti-label-export-table doesn't have units, it produces a text version of the correspondence between integers and label names/colors in a dlabel file (which contains integers labeling each vertex/voxel, and not any sort of data that has units).

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Wednesday, June 9, 2021 3:27 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 15, 2021, 4:54:38 PM6/15/21
to hcp-...@humanconnectome.org
Hi Tim,

For -cifti-create-dense-from-template, I wanted to check if I am using the temporal labels, would I use CORTEX_LEFT and CORTEX_RIGHT for -metric structure? I didn't see an option for temporal structures in the list.

Since -cifti-convert creates just a list of numbers, do the numbers correspond to order of labels in the label file?

Is there a way to get curvature data from the -cifti-parcellate file?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Wednesday, June 9, 2021 4:27 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 15, 2021, 7:06:16 PM6/15/21
to hcp-...@humanconnectome.org
Yes, use CORTEX_LEFT and CORTEX_RIGHT for surface data, we don't divide that into pieces in cifti format.

The order of the parcels in the -cifti-parcellate output is the same as the sorted label key integers, which is the order -*-export-table uses also.  -cifti-convert will keep the order of whatever the input file was.

Curvature data is created from surface files, you can put it into cifti format and process it the same way as any other data.  However, curvature is entirely based on folding patterns, which are loosely connected to functional areas, so we only use it to generate a "corrected" thickness (the basic thickness measurement shows changes due to curvature, so we regress them out).

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Tuesday, June 15, 2021 3:54 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 18, 2021, 2:31:04 PM6/18/21
to hcp-...@humanconnectome.org
Hi Tim,

Can a dscalar file be either thickness or volume? If I used *_corr_thickness.dscalar.nii as input for -create-dense-from-template, would this create a dscalar for thickness? What would I use to make a dscalar for volume that I can be used in -cifti-parcellate? I used the same dscalar file in -cifti-parcellate but thought if I used MEAN it would output thickness and SUM would output volume.

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Tuesday, June 15, 2021 4:54 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 18, 2021, 3:41:59 PM6/18/21
to hcp-...@humanconnectome.org
Yes, you can put any type of data you want into dscalar format, the pipelines make dozens of dscalar files per subject, containing many different measures (myelin, task activation, thickness, sulcal depth, etc) and a few processing options (which registration, what mesh density).  It is like nifti, just a way to store values.  You may want to read through wb_command -cifti-help, but the important bit is that cifti is really just a matrix of values, with metadata that describes what each index means (which vertex, the name of the map).  It does not contain geometry or surface coordinates.

The -surface-wedge-volume command will give you metric files containing cortical volume per vertex.

-cifti-parcellate can only operate on the values provided in the input file(s).  It doesn't understand anything about geometry, it is all just numbers.  However, it would not make much sense to sum values of thickness across vertices, and likewise it would not make much sense to take the mean of the per-vertex volume, which is why I said to use different reduction types on different input data (you still have to give it the different input data by giving it different input files).  wb_command has no built-in logic to tell you if you are doing something that looks useless, think of it like a calculator.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 18, 2021 1:30 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 18, 2021, 4:10:36 PM6/18/21
to hcp-...@humanconnectome.org
Hi Tim,

If I wanted to get thickness values from -cifti-parcellate, would I just use the *_corr_thickness.dscalar.nii as input without using -create-dense-from-template?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 18, 2021 2:30 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 18, 2021, 4:20:16 PM6/18/21
to hcp-...@humanconnectome.org
Yes, that will work, because it is already in cifti format.  Note that "corr_thickness" has had the curvature measure regressed out of the raw thickness measurement.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 18, 2021 3:10 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

Hobert, Leah E.

unread,
Jun 20, 2021, 7:38:15 PM6/20/21
to hcp-...@humanconnectome.org
Hi Tim,

Does that mean I could also use this to get curvature? Which -method would be used for -cifti-parcellate? Or would I have to use the *_curvature.dscalar.nii with -cifti-parcellate or *_left_curvature.shape.gii wirh -cifti-create-dense-from-template?

Best,
Leah
From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Friday, June 18, 2021 4:10 PM

Coalson, Timothy Scott (S&T-Student)

unread,
Jun 21, 2021, 5:07:54 PM6/21/21
to hcp-...@humanconnectome.org
The output of -cifti-parcellate is always the same measure as the input - if you want parcellated curvature to be the output, the input needs to be "dense" (the D in dscalar) curvature.

I don't know that parcel-level curvature would be particularly informative.  To me, the sum of the curvature of the vertices seems even less likely to be useful than the mean.  I think the curvature measure is signed (and there is more than one measure of curvature), so you would also need to decide whether you wanted positive and negative curvature to cancel out, etc.

Tim

From: Hobert, Leah E. <LHO...@PARTNERS.ORG>
Sent: Sunday, June 20, 2021 6:38 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Re: dHCP statistics
 
--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
Reply all
Reply to author
Forward
0 new messages