Dual Volume - Volume Rendering

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Dual Volume - Volume Rendering

devakumar-3
Hi
Volume Rendering with Single volume of data works fine. But with dual volume, the algorithm "NCI GPU Ray Casting (Multi-Volume) doesn't work. I have tried in two systems.
System 1:
Slicer 3.6 RC1 (Slicer 3 3.5.2010-05-09) Stable Release
Windows XP 32 bit
GeForce GTX 285 (1GB video memory)
4GB of RAM
Problem: Severe artifact in rendered image (not able to recognize the image)

System 2:
Slicer 3 3.5.2010-05-17 Snapshot
Windows XP 64 bit system
Quadro FX 5800 card (4GB video memory) and  also.
48GB RAM
Problem: 3D Display window is not updated. Ray casting is done indefinitely.

I have tried with todays nightly build and last two days of nightly build also. I cannot use CPU ray casting as it supports only one volume.

Appreciate your help.

Thanks
Kumar
----------------------------------------
Devakumar Devadhas, Ph.D
Postdoc Fellow
Department of Radiology and Molecular Biology
Thomas Jefferson University
Philadelphia
PA-19107

 
_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject
Reply | Threaded
Open this post in threaded view
|

Re: Dual Volume - Volume Rendering

Yanling Liu
Hello Kumar,

It is interesting that you are using CT and PET volumes with different size and spacing. The multi-volume module is currently under development and the initial goal was to compare pre and post treatment images. You could imagine in this case both bg and fg images will have same spacing, size, etc. While I am excited for the new requirement to handle images with different modality, size, and spacing, I'll need some time to do coding. Some screen shots will also help me to better understand it. At the meanwhile, you may able to render images if you can do some padding and resampling to PET dataset.

Are you able to provide me some sample CT/PET dataset pair so I can test multi-volume module in development?

Yanling

From: Devakumar Devadhas [[hidden email]]
Sent: Tuesday, June 01, 2010 3:22 PM
To: Liu, Yanling (NIH/NCI) [C]; [hidden email]; [hidden email]
Subject: RE: Dual Volume - Volume Rendering

Hi

Thanks for your response. The technique I used is Composite with Shading. I have tried all other techniques which you have provided.

Multi-Volume GPU Ray casting doesn't have Depth peeling option. I have tried Bg/Fg Ratio. The size of the CT dataset is 512x 512 x 431. The size of the PET dataset is 128 x 128 x 344. I tried with the same size assuming that your GPU ray casting module may resample the data to half the size. I resampled the data and tried also. Neither worked. Ofcourse image spacing is not same in both datasets. It may be very hard to have same image spacing unless both are CT datasets and acquired in the same machine. I have the latest NVIDIA drivers installed in the system.

Thanks

Kumar

-----------------------------------------------------
Hello Kumar,

Could you please provide me more details for the problem? I would like to know which techniques you have selected for bg and fg channel. Did you change parameters such as distance color blending or volumetric depth peeling? What is the size of your data set? Do bg and fg volumes have same origin, size and spacing? For now the multi-volume rendering mapper is still under development and won't handle bg and fg volumes with different size, spacing and/or origin.

Also please make sure the latest Nvidia driver is correctly installed in your system.


On Mon, May 31, 2010 at 8:47 PM, Devakumar Devadhas <[hidden email]> wrote:
Hi
Volume Rendering with Single volume of data works fine. But with dual volume, the algorithm "NCI GPU Ray Casting (Multi-Volume) doesn't work. I have tried in two systems.
System 1:
Slicer 3.6 RC1 (Slicer 3 3.5.2010-05-09) Stable Release
Windows XP 32 bit
GeForce GTX 285 (1GB video memory)
4GB of RAM
Problem: Severe artifact in rendered image (not able to recognize the image)

System 2:
Slicer 3 3.5.2010-05-17 Snapshot
Windows XP 64 bit system
Quadro FX 5800 card (4GB video memory) and  also.
48GB RAM
Problem: 3D Display window is not updated. Ray casting is done indefinitely.

I have tried with todays nightly build and last two days of nightly build also. I cannot use CPU ray casting as it supports only one volume.

Appreciate your help.

Thanks
Kumar
----------------------------------------
Devakumar Devadhas, Ph.D
Postdoc Fellow
Department of Radiology and Molecular Biology
Thomas Jefferson University
Philadelphia
PA-19107


_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject


_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject
Reply | Threaded
Open this post in threaded view
|

Re: Dual Volume - Volume Rendering

devakumar-3
In reply to this post by devakumar-3
Hi Yanling
Here is the link for the PET-CT dataset. http://pubimage.hcuge.ch:8080/DATA/PROSTATIX.zip
This is from University of Geneva website. If I have to send my own data I should anonimize the data, but this is already anonomized and readily available.

Or you can go to http://pubimage.hcuge.ch:8080/ and choose your own PET-CT dataset. Most of the datasets in this site are CT and there are few PET-CT dataaset also. I will send the screen shots later.

Could you please tell me some way of padding. I am familiar with resampling, but I have not done padding.

Thanks
Kumar
---------------------------------------------------------
Hello Kumar,

It is interesting that you are using CT and PET volumes with different size and spacing. The multi-volume module is currently under development and the initial goal was to compare pre and post treatment images. You could imagine in this case both bg and fg images will have same spacing, size, etc. While I am excited for the new requirement to handle images with different modality, size, and spacing, I'll need some time to do coding. Some screen shots will also help me to better understand it. At the meanwhile, you may able to render images if you can do some padding and resampling to PET dataset.

Are you able to provide me some sample CT/PET dataset pair so I can test multi-volume module in development?

Yanling
_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject
Reply | Threaded
Open this post in threaded view
|

Re: Dual Volume - Volume Rendering

Yanling Liu
Hello Kumar,

Please check out the image showing PET-CT visualization in Slicer. http://www.slicer.org/slicerWiki/images/1/1f/PET-CT.png

Here are steps I performed for PET-CT visualization:

1. downsample CT volume to 256x256x369 (512x512x738 is too large in multi-volume mode even for a 1GB video card)
2. use Slicer Resample Scalar/Vector/DWI Image module to resample PET volume to be same size, origin, spacing with the down sampled CT volume
3. In volume rendering module, set resampled PET volume as bg volume and down sampled CT volume as fg volume (reversed order may cause Slicer freezing, I need to figure out why it happens)
4. change rendering technique for both bg and fg to Gradient Magnitude Opacity Modulation
5. adjust transfer function and bg/fg ratio slider for your image.

In the sample image above, gray colors are from PET volume, all non-gray colors are from CT volume.

Yanling

On Thu, Jun 3, 2010 at 11:11 AM, Devakumar Devadhas <[hidden email]> wrote:
Hi Yanling
Here is the link for the PET-CT dataset. http://pubimage.hcuge.ch:8080/DATA/PROSTATIX.zip
This is from University of Geneva website. If I have to send my own data I should anonimize the data, but this is already anonomized and readily available.

Or you can go to http://pubimage.hcuge.ch:8080/ and choose your own PET-CT dataset. Most of the datasets in this site are CT and there are few PET-CT dataaset also. I will send the screen shots later.

Could you please tell me some way of padding. I am familiar with resampling, but I have not done padding.

Thanks
Kumar
---------------------------------------------------------
Hello Kumar,

It is interesting that you are using CT and PET volumes with different size and spacing. The multi-volume module is currently under development and the initial goal was to compare pre and post treatment images. You could imagine in this case both bg and fg images will have same spacing, size, etc. While I am excited for the new requirement to handle images with different modality, size, and spacing, I'll need some time to do coding. Some screen shots will also help me to better understand it. At the meanwhile, you may able to render images if you can do some padding and resampling to PET dataset.

Are you able to provide me some sample CT/PET dataset pair so I can test multi-volume module in development?

Yanling


_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject