Cortical surface toolbox - surface analysis pipeline

Questions about BrainVisa usage and installation

Moderators: denghien, riviere

aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Hello,

I would like to know the volume of different gyri.
For that I ran the parametrisation of cortical surface pipeline. The operation succeeded and all seems to be good.

After that, to have the volume, I tried to launch the surface analysis pipeline. But I don't know how to fill in the field "contrast". it seems it has to be a "string" (letters) but I don't understand what is it...
What should I write here?

Thank you very much,

Aurélie.
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Hello,

I'm sorry to post again, but it would be really great if I could have a reply... :)

Thank you very much,
aurely1310
User avatar
Olivier Coulon
Posts: 176
Joined: Fri Feb 27, 2004 11:48 am
Location: MeCA research group, Institut de Neurosciences de La Timone, Marseille, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by Olivier Coulon »

Hello Aurely,
For that I ran the parametrisation of cortical surface pipeline. The operation succeeded and all seems to be good.
Well done !
After that, to have the volume, I tried to launch the surface analysis pipeline. But I don't know how to fill in the field "contrast". it seems it has to be a "string" (letters) but I don't understand what is it...
The Surface analysis pipeline is for processing functional data at the group level on the cortical surface. So it's not what you're looking for.
At the end of the parameterization pipelmine, a texture of gyris is generated. That might be what you want to use for your problem. But I don't know what exactly you mean by the 'volume' of gyri. Is it the gery matter volume ? If so, it is not implemented yet, but will be. You can try to run another gyral parcellation process, though : the one developed by Arnaud Cachia, in the 'old' folder of the cortical surface toolbox. I think he provides the gray matter volume for each gyrus.

Does it help ?

Olivier
Olivier Coulon
Institut de Neurosciences de La Timone,
Aix-Marseille Université,
Marseille, france
https://meca-brain.org
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Hello,

Thank you for your reply.
It would be ideally great.... It was exactly what I wanted to do but...

i tried to launch the "old" "gyri parcellation" but nothing happened : After 1s, I've got the message "operation succeded (1s)".

So it doesn't seem normal!!!

How can you explain that?

Thank you very much,
aurely1310.
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by riviere »

Hi,
I have tried the parcellation process today.
Did you input *labelled* cortical graphs ? The process code is somewhat dirty, there are imprecise input types and broken parameters links. I have spent several hours fixing it this evening and it should behave better now, but if all parameters were (sometimes manually) set, the process seems to work.
Remember to select the Parcellation parameter as 'Cortical Ribbon' if you want a volumic parcellation where you will be able to measure gyral volumes.
Denis
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Hi,

I tried to fill in manually each field ("old process"), but it was impossible to fill in "left_white_sulci", "right_white_sulci", "left_white_giry_volume" and "right_white_giry_volume".
May be there is an operation that I have to do before to launch this process?

Thanks,
Aurely1310
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by riviere »

Hi,
That's precisely the problem I ran into yesterday. But these parameters are output files, you can enter whatever filename you want (provided the file extension is OK, it may be a non-existing file, but in an existing directory, for instance /tmp/left_white_sulci.tex, /tmp/right_white_sulci.tex, /tmp/left_white_giry_volume.nii and /tmp/right_white_giry_volume.nii.
It will be fixed in the next version 4.0 which is just about to be released... (if we don't find other critical bugs at the last minute...)
Denis
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Thank you for your reply :

I had these messages :
Warning: wrong filename_base in graph, trying to fix it
Not able to close the projected connnected component, but I continue.
Not able to close the projected connnected component, but I continue.
Warning: wrong filename_base in graph, trying to fix it
The point can't be projected (the projected node is too far from the the initial voxel
The point can't be projected (the projected node is too far from the the initial voxel
The point can't be projected (the projected node is too far from the the initial voxel
The point can't be projected (the projected node is too far from the the initial voxel
The point can't be projected (the projected node is too far from the the initial voxel
The point can't be projected (the projected node is too far from the the initial voxel
The point can't be projected (the projected node is too far from the the initial voxel
Not able to close the projected connnected component, but I continue.
Not able to close the projected connnected component, but I continue.


But globally the algorithm run correctly now.

So, now how can I obtain gyral volumes?
(the button "eye/show" of left_white_giry_volume and right_white_giry_volume is not activated...)

Thank you very much,

aurely1310
User avatar
Olivier Coulon
Posts: 176
Joined: Fri Feb 27, 2004 11:48 am
Location: MeCA research group, Institut de Neurosciences de La Timone, Marseille, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by Olivier Coulon »

The warnings are alright, they are just leftovers of old debug information.

To get grey Matter volume, I think you need to make sure that the 'Parcellation' field is set to 'Cortical Ribbon'. After that I am not sure because I did not write this process but I think that one of the output arguments is a volume representing the cortical ribbon with different values for each gyrus. Then it is a question of counting voxels for each gyrus. I don't know if there is any program for this.

Olivier
Olivier Coulon
Institut de Neurosciences de La Timone,
Aix-Marseille Université,
Marseille, france
https://meca-brain.org
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by riviere »

Hi,
To obtain gyral volumes, there are several ways:
- in the output gyri graphs (the .arg files), there is a "size" property and a "surface_area" property in each gyrus entry: size is the volume of the grey matter voxels in the gyrus, and surface_area the surface of the gyrus (actually I don't remember if it's the external or internal surface, I guess external but I'm not sure).
- to extract them automatically in table-like files and on several subjects, you can use the morphometry statistics process of brainvisa, in the morphometry toolbox. Then you should use the gyrusModel or the roiModel model file, and the gyri.hie normenclature. The gyrusModel will also provide the mean thickness of the cortex on each vertex (it's simply the volume divided by the surface). Actually there is a bug in the gyrus model, I've just fixed it, so it may not work on your data. It should be OK in the next release...
- otherwise you can also use the AimsVoistat command on either the gyri graph or the gyri volume which was outpout from the parcellation process. But this command was designed to calculate stats on additional functional data, so it will require another volume (even if you don't need it, you can also provide the gyri volume).
- or the AimsRoiFeatures command will also do the volumes calculation on ether the graph or volumes of gyri, its result may be viewed by the BrainVisa process "BrainVisa Show Scalar features" in tools/viewers.

Denis
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Hello,

I tried do do what you said but I met some problems for each way you adviced...

1. I can see the field with the "giry graph" buut when I want to open it, it is impossible. How can I open it? Do I have to launch it with anatomist separetely? (the "show" button doesn't work...)

2. The process seems work! But the results seems to be wrong : "Mathilde F.C.L.r.retroC.tr. left 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.975435" . I think that's your problem?

3 & 4. i don't understand how to use these process...

Thank you very much,
Aurélie.
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by riviere »

Hi,

1. There is no viewer icon on the morphometry process graph input because the input type is a list of graphs (and we haven't implemented the viewer), but you can use directly the "anatomist show gyrus graph" viewer process in cortical surface/viewers. This will use Anatomist to show the graph, and then you need to see its structure and attributes: in anatomist you should put the graph object in a browser window. Otherwise loading the .arg directly in anatomist is OK, of course.
The .arg file is a text file, so you can also use a text editor to see it.

2. the result is wrong because it's the wrong model which is being used: the default model is a sulci model, it will compute information which is dedicated to sulci, which doesn't match here. You should change the model parameter and use either the "gyrusModel" or the "roiModel" models, as I explained in the previous message. You also have to change the nomenclature to the "gyri" nomenclature. These two parameters will not be set automatically (unfortynately).

3/4. These are not BrainVisa processes, but commandlines: programs that you need to run by hand, from a terminal. All commands are in the bin/ directory of the BrainVisa package, and running them with the -h option will output a little help about how to use them. Or you can look at http://brainvisa.info/doc/documents/shf ... mands.html.

Denis

PS: by the way, brainvisa 4.0 is finally released, the above bugs are fixed...
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Hello,

I tried to follow your advises, but :

1. Even if, I succeded to show the gyrus graph, it seems there is a problem with the display. At the beginning, the image seems to be good, but when we want to rework it (for example to make a zoom) the image looks very bad...!

First image :
show_gyrus.png
show_gyrus.png (68.45 KiB) Viewed 10144 times
Next :
show_gyrus_pb.png
show_gyrus_pb.png (58.02 KiB) Viewed 10141 times

2. I tried to change parameters, but it is impossible to select the model...I can just change the "filename variable"... :(

Then, I will try to use the 3rd and 4th points.

Oh, I just downloaded the new version of brainvisa. For the moment I cannot see changments (except the images of anatomist) but I'm sure that there is a lot of bugs which don't appear. (Probably also all the bugs on windows?)
Does this version help me for what I want to do?

Another question, I wrote a topic about my problems on fiber tracking but had no answer...could you reply on the other topic please? It would really be great...! :)

Thanks a lot,
Aurélie.
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Cortical surface toolbox - surface analysis pipeline

Post by riviere »

Hi Aurélie,

1. there are 2 problems in the display actually:
a. the gyrus graph referential is wrong (it's not set when the graph is created). This is fixed in 4.0.
b. the meshes appear weird (a problem of polygons inside/outside notion or normals orientation). I didn't notice it before the release of version 4.0 so it's not fixed. I have to look at it. But it seems that it's only a display problem, the results and voxels are OK I think. You can display the voxels (which are OK): in anatomist, open the "graph parameters" settings (in the main window menu) and select "display facets".
Anywayn, it shouldn't prevent you from using a browser window to see the graph structure and properties.

2. Don't you see several models when clicking on the green database button on the model parameter ? Even after clicking "reset" in the database box ? You should see 6 of them in BV 3.2.1 (and 8 in BV 4.0):
models.jpg
models.jpg (85.68 KiB) Viewed 10129 times
BV 4.0 fixes a number of bugs, but maybe not all of them - to be honest, strange behaviours on Windows that I could not reproduce have not been worked on. But it integrates some fixes in the gyral parcellation, and gyri morphometry model, that should be useful for you.

For you fibers tracking problem, well, I don't really know, I'm not the specialist of fibers tracking; Yann is quite busy and not here at the moment, and other fiber tracking people just don't read the forum I guess...

Denis
aurely1310
Posts: 38
Joined: Tue Apr 06, 2010 10:22 am

Re: Cortical surface toolbox - surface analysis pipeline

Post by aurely1310 »

Ok, the 2nd way seems to work!!! :)

I just have this message :

Morphometry statistics running
config : /tmp/bv_293629.cfg
Gyrus cliques should have 1 vertex, 2 found here. clique label: Temporal-Middle_left Gyrus cliques should have 1 vertex, 2 found here. clique label: Frontal-Middle_left


Process Morphometry statistics 1 finished on 2010/06/18 15:14 (7 seconds)


So, some files don't work... :
1. some files don't have a name (called "Giry_gyrus8_left.dat")
2. Some files seem very strange!!! : "subject label side surface_area size thickness gyrus_descriptor_output
Mathilde Post-Central left 0 11487 inf 0"

And what means "size" ? the volume? in voxels?


I'll post again for the 1st case. I'll try now...

Thanks, Aurélie.
Post Reply