razvanmarinescu / brain-coloring Goto Github PK

I am running Ubuntu virtual machine on windows. When I use make after following the YouTube tutorial I get the following error
The command seems to run fine but then right at the end as it is trying to save the image a syntax error occurs.

Saved: 'C:\output\mice_output\top_Image_1.png'
 Time: 00:02.83 (Saving: 00:00.14)

The syntax of the command is incorrect.


Blender quit
(brainpainterEnv) wintermute@LAPTOP-6F6LEGSI:/mnt/c/Users/angus/Desktop/brain-coloring$

My windows is x64 bit. I'm running the script inside a conda environment with the appropriate libraries (such as the specified version of pandas).

Thank you for the website update!
I have tried with the new Destrieux template and now it is working properly. Thanks!

I have an additional question about the input range.
If I want to plot some ROI values range from 0-max (max can be larger than 10, say 12), rather than just 0-3 or 0-4, and I still want to use 3 colors (say yellow/orange/red), is that I just need to edit the latex file for color bar?

For example, I just need to change those latex syntax:

\draw (0,0) -- (0.5,0);\node[inner sep=0] (corr_text) at (1.6,0.0) {normal};
   
\draw (0,2) -- (0.5,2);\node[inner sep=0] (corr_text) at (1.6,2.0) {1-sigma};
   
\draw (0,4) -- (0.5,4);\node[inner sep=0] (corr_text) at (1.6,4.0) {2-sigma};
   
\draw (0,6) -- (0.5,6);\node[inner sep=0] (corr_text) at (1.6,6.0) {3-sigma};

To:

\draw (0,0) -- (0.5,0);\node[inner sep=0] (corr_text) at (1.6,0.0) {0};
   
\draw (0,2) -- (0.5,2);\node[inner sep=0] (corr_text) at (1.6,2.0) {4};
   
\draw (0,4) -- (0.5,4);\node[inner sep=0] (corr_text) at (1.6,4.0) {8};
   
\draw (0,6) -- (0.5,6);\node[inner sep=0] (corr_text) at (1.6,6.0) {12};

Am I correct?

Thank you!

Hi,

Currently it seems the camera only shows the right side of brain (and only 1/2 of the regions are needed in the csv file), but at times it would be useful to show the differences between right vs left brain (e.g. non-symmetrical diseases). Is there a supported way to do this, currently?

Thanks!

Hello,

I tried again the website for plotting with the new Destrieux atlas template (consistent with the paper format), but it still show errors as below:

The following regions are missing from the input .csv file: ['Thalamus-Proper', 'Lateral-Ventricle', 'Cerebellum-Cortex', 'Putamen', 'VentralDC', 'Amygdala', 'Inf-Lat-Vent', 'Hippocampus', 'Pallidum', 'Caudate', 'Accumbens-area', 'Cerebellum-White-Matter']

Make sure the correct atlas is used, and double check the example template corresponding to that atlas. Note that, currently, all regions from the template need to be assigned a number mapping to a color.

Could you please kindly help me with this problem?

Thanks!

Hi @vmallela0,

How hard would it be to add a mouse brain atlas? Specifically, I'm thinking about the Dorr and DSURQE atlases (https://wiki.mouseimaging.ca/display/MICePub/Mouse+Brain+Atlases)

Let me know what you think.

I have another question about this tool.

I just noticed that the order of ROIs name in the Destrieux atlas template csv file is slightly different from the order in the original paper where the Destrieux atlas was proposed.
"Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature"

I have an array of 148 ROIs values which is consistent with the order in the orginal paper, but the template has a different order of ROI.

Can I modify the order of ROIs name in the template csv file and fill in the corresponding values for each ROI?
This means that the tool will match the values with the brain regions according to the ROIs name.

Or, the BrainPainter tool only takes the values by that order specified in the template csv file?

Thanks!

after i call make, i get this error (this is on MacOS Big Sur, version 11.2.3):

configFile=config.py /Applications/Blender/blender.app/Contents/MacOS/blender --background --python blendCreateSnapshot.py
/bin/sh: 1: /Applications/Blender/blender.app/Contents/MacOS/blender: not found
Makefile:2: recipe for target 'all' failed
make: *** [all] Error 127

Is there possibly a dependency/step missing in the instructions for installation using Docker?

Thanks in advance for your help!

Hi, I'm trying to follow the steps for testing out brain-coloring. Should there be a data/pcaCover.csv file here somewhere? Sorry if I'm missing something obvious...

I am having a problem with the color bar.
I cannot make the color bar range from 0-3 using the browser.

Is possible to adjust the range of color bar from the browser?
Or, I have to install the docker to adjust it?

First of all: great tool! I really like the simplicity with which you can generate high quality visualisations.

Now to a small bug I might have found. I'm using the Destrieux atlas and would like to visualise some measurements for the different ROIs. I'm not particularly interested in the subcortical regions at this stage, so I had left them out -- thankfully, the web tool raised an error suggesting that some expected columns were missing from the .csv I uploaded. I added some dummy values for these regions (based on the suggested columns highlighted as missing) and the image generation seemed to kick off without any issues.
However, at the end of the process I got a warning that some images could not be generated and the logs weren't too helpful as they seemed to contain mostly lines indicating successful imports, suddenly an incomplete line and ending in Blender quit (see below).

...
Successfully imported 'models/Destrieux_atlas_pial/rh.pial.Destrieux.G_insular_short.ply' in 0.175 sec

Successfully imported 'models/Destrieux_atlas_pial/rh.pial.Destrieux.G_pariet_inf-Supramar.ply' i

Successfu

Successfully imported 'models/Destrieux_atlas_pial/rh.pial.Destrieux.G_and_S_frontomargin.ply' in 0.244 sec

Successfully imported 'models/Destrieux_atlas_pial/rh.pial.Destrieux.G_and_S_cingul-Ant.ply' in 0.485 sec
...
-------------generated/F8FA9DF9DE524ED1/F8FA9DF9DE524ED1_wrong-example.csv-------------

Blender quit

I double checked my .csv against the example you include for download on the website. The only difference I could find is that my data does not contain the [Left-Unknown, Right-Unknown] columns but I am pretty sure these weren't highlighted/included in the initial error message when actual ROIs were missing. Once I added these two columns with some dummy values to my .csv everything worked fine. This seems to be a really minor issue but it could trip up people as the initial error message indicating missing columns does not include this.

Feel free to ignore this in case I mixed up something when trying to reproduce the error -- this could well have happened. If this is an actual issue, I'm happy to look through the source and fix it once I have the time for it. Otherwise, any help/input from your side would be appreciated. Cheers!

Dear Mr Razvan Marinescu:
I am writing to ask whether you could help me. When loading .csv into BrainPainter, an error occurs. "The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application."

Dear team,

Thank you for this wonderful software! We are using this software to map pathology that are scaled from 0-1 or non-scaled values which are heavily skewed. Our values are heavily skewed in the range of 3.5-5, hence no matter what color we choose it ends up being a very subtle shade between last two colors in the range of pathology [0-5]. Same thing happens if I scale these values to 0-1. Is there a way I can show a scaled distribution of pathology values with hot color map or any other map that can show clear distinction of color on the glass brain? As in if I want to map scaled values that are in the range of 0-1 with say hot colormap (black -> red -> orange -> yellow), is it possible to do that? When I tried doing this, values were mapped to black and/or red discarding further interpolation from red to yellow.

Thank you,
Sneha

hello,
i am trying to generate some figure in your website but got some error that

The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application.

thanks a lot for fixing

Dear Razvan,
First of all congrats for such a great tool! :) I hope my question is not redundant, but i really have issues with the output_folder. I don't understand why for the template data the output "temp" works, but when i put my own data sheet (metadata is according to the DK atlas as seen in the template data), neither the temp or a chosen directory works. I'm using the web-based tool.

Thanks already,
Moritz

Hello,

I noticed that the Destrieux template in the github/input folder has been updated as the paper format.

May I ask if the website has been updated to support this new template?
If so, I just noticed that the template on the website was still the old one.

Thank you!

Hello,

I just have a quick question regarding the input range.

I just wanted to confirm that if BrainPainter only takes input values range from 0 - 3?
So, it means that BrainPainter cannot handle input range from 0 - m, where m represents any positive integer?

Thanks!

Hi @razvanmarinescu,

I'm trying to run BrainPainter off the website and keep getting an 500 internal server error. Can you confirm that it is up and running?

Thanks

Hello there,

Currently, the BrainPainter webpage just stop working properly. It keeps running but cannot show completed.

Can you take of a look at it?

Thanks!

Hi,
In our lab we are really keen on using your toolbox to plot some data we have using some custom parcellations. We have created all the files (i.e., the .ply files) for all our parcellations and would like to know how we can edit your code to use our parcellations for plotting, if this is at all possible.

Thanks in advance.

Hello,

For some reason on the last version 59f2352, the resulting images are different every time I launch the make. I'm using blender 2.79 and Ubuntu 18

How to reproduce:

1. Run make first time. Some output will appear.
2. Rename pictures and keep them in the same folder.
3. Run make second time.
4. Visually compare images.

Here I attach some images for mice_template.csv obtained by above mentioned procedure:

For DK_template it is the same - I can't paint specific segment because it maps to wrong segments every time I launch make. Looks like the inner color mapping constantly misses the elements.

Hi @razvanmarinescu ,

Thanks a lot for the great software.

The top/bottom view of brain_type='inflated' I am getting looks wrong.

I am using your repository to display feature importance of different brain regions. I want to convey postive and negative values on the same color gradient I want to make sure this is the correct way to do so.

So I will use the following change to the config.py file

# NOTE: you can add more than 4 colours (as many as you want), in which case you should increase the range of the input biomarkers.
COLORS_RGB = [(0,0,1), (0,0,0.5),(1,1,1),(0.5,0,0),(1,0,0)] # Blue ->White-> Red

My starting scale is from -4 to 4 so I would convert it to be between 0-6 (so a feature importance of 0 in my orginal data would become 3 (corresponding to white).

Lastly How can I get the color gradient bar legend to display the values as they relate to my orginal data? So it would be marked as ranging from -4 to positive 4 rather than 0-5?

Hello,

I am trying to using the website for plotting, but I am getting error message:
Internal Server Error
The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application.

Thanks!

Hi there,

I am using the BrainPainter website to generate images to display significant DK ROIs.
However, I noticed that the inner right hemisphere view shows the subcortical structures.

Is there a way to remove the subcortical structures from being displayed by using the BrainPainter website?

I tried deleting the columns for the subcortical structures from the DK_template.csv that was provided on the website but this does not work and will just give an error message that those regions are missing from the csv file.

Thank you!

Is there any chance you have the .ply files for the substructures of the brain? for example, within the lh.pial.Mice.Striatum.ply file, having subvolumes for the caudate and putamen. In the infralimbic cortex, subvolumes for the dlPFC, mPFC, etc.

I know that all these components are annotated in the Allen atlas. but it looks like .ply files in this repository only exist for the higher level brain structures. Having access to the lower level structures in Blender would be super helpful for scientific animation purposes.

Is there a way to add a view that displays the insular cortex with the temporal lobe removed? If I colour it now it is seen only as a spec of colour on the lateral view.

Sorry I still have a question regarding the range of input ROI values.

For example, if I want to plot some ROI values ranged from 0-12, rather than the default setting of 0-3, how should I modify the source code for this?

Currently, I am able to plot for the values of 0-3, but I need to do a minmax normalization to have my data fit for range of 0-3.

But now, I would like to plot my ROI values on its original scale. How could I plot the input values from 0-12 or 0-max?

Thanks!