shapeworks-users

Text archives Help


[shapeworks-users] DeepSSM


Chronological Thread 
  • From: Claudia Companioni-Brito <ccompanionib@gmail.com>
  • To: Alan Morris <amorris@sci.utah.edu>, shapeworks-users@sci.utah.edu
  • Subject: [shapeworks-users] DeepSSM
  • Date: Wed, 4 Nov 2020 18:18:53 +0100
  • Ironport-sdr: 6Uxud8nxFBQRKwF7P0umJF90RtzFSclfCQsZ3+IbjZB8200PenFSO3494YLZjBhWHzg/LeZ6Db UftM0mvKzpXAbF/0EAKvf2RK75yn06RNFmZZYKEBnjrA4PvQTmvRkGA7qOidQChy9ag1efBSn1 gvc64dcAwBwrA0RgoOfJnG0odRb3tawnspCRk7rYTnJKkoVF3Kbj96z4ZFalffKpIJJybW+M8y +eDh72iaAJC6TnMhz3IsV0Mg3qmBeoqINLAVBBNZXANt4KptRDZR5uEVPeOh26ZnaFyk5l1Ke0 riE=

Hello Alan,
I'm trying the new DeepSSM but every time I get this error: RuntimeError: storage is wrong size: 0 is expected to be 2
Could you help me?

Thank you,
Best, Claudia

On Tue, 19 Nov 2019 at 11:59, Claudia Companioni-Brito <ccompanionib@gmail.com> wrote:
Hi Alan,

I tried IsotropicResampling with 0.1, 0.05 voxel size.

Best,
Claudia



On Mon, 18 Nov 2019 at 20:51, Alan Morris <amorris@sci.utah.edu> wrote:
Claudia,

I will take a a look.  What voxel size did you resample to?  0.5?

Thanks,
Alan


On Mon, Nov 18, 2019 at 12:08 PM Claudia Companioni-Brito <ccompanionib@gmail.com> wrote:
Hello Alan,

Thank you so much, It is working well now after grooming but I am having some troubles during the optimization step ( Studio and ShapeWorksRun crashes). Please, see the attached images of the error.

Best,
Claudia  
 

On Sat, 9 Nov 2019 at 17:16, Alan Morris <amorris@sci.utah.edu> wrote:
Hi Claudia,

I think your data needs a combination of a smaller voxel size (e.g. 0.05) and/or the "TopologyPreservingSmoothing" tool.  This is my first time using this tool, but I did some experimenting and I think we can get this to work.  See the attached image, I also lowered the Blur sigma to 0.20.

Alan


On Thu, Nov 7, 2019 at 7:34 PM Shireen Elhabian <shireen@sci.utah.edu> wrote:
Hi Alan,

Would you mind looking into this? Attached are sample shapes.

thanks and best regards
Shireen


------------------------------------------
Shireen Elhabian, M.Sc., Ph.D.
Research Assistant Professor
School of Computing
Scientific Computing and Imaging Institute
University of Utah
WEB 2815
72 Central Campus Drive, Salt Lake City, UT, 84112
Phone (801) 587-3206
Fax (801) 585-6513
Home Page: http://www.sci.utah.edu/~shireen




On Oct 28, 2019, at 2:34 PM, Claudia Companioni-Brito <ccompanionib@gmail.com> wrote:

Hello Shireen,

Thank you very much!

Following your advice, I tried to adjust the voxel spacing (to 0.1 and less) for IsotropicResampling and it was good, but I still have problems after applyDistanceTransforms. I have changed some parameters in this step (antialiasIterations, smoothingIterations, alpha, beta) but the shapes keep losing its original form.
 
Best,
Claudia

On Wed, 23 Oct 2019 at 20:17, Shireen Elhabian <shireen@sci.utah.edu> wrote:
Hi Claudia,

The shapes disappear because these are so thin structures that do not have enough pixel-wise spatial support.  What voxel size are you using for the “applyIsotropicResampling” step? The segmentations voxel spacing is 0.1, if you are using a larger voxel spacing, this would make the shapes disappear. You could also resample it with less than 0.1 space to get a better resolution. Also, these segmentations have enough background voxels, you might want to bypass the padding step. 



<snapshot0001.png>
thanks and best regards
Shireen


------------------------------------------
Shireen Elhabian, M.Sc., Ph.D.
Research Assistant Professor
School of Computing
Scientific Computing and Imaging Institute
University of Utah
WEB 2815
72 Central Campus Drive, Salt Lake City, UT, 84112
Phone (801) 587-3206
Fax (801) 585-6513
Home Page: http://www.sci.utah.edu/~shireen

On Oct 22, 2019, at 5:10 AM, Claudia Companioni-Brito <ccompanionib@gmail.com> wrote:

Dear Shireen Elhabian,

Thank you very much for your quick response.
Attached is a .zip file with 5 shape samples I've been testing with Shapeworks.

King regards,
Claudia

On Fri, 18 Oct 2019 at 19:21, Shireen Elhabian <shireen@sci.utah.edu> wrote:
Hi,

Would you mind sharing with us some shape samples (e.g. 3 segmentations) that you are using as input to Studio to help us troubleshoot this?

thanks and best regards
Shireen


------------------------------------------
Shireen Elhabian, M.Sc., Ph.D.
Research Assistant Professor
School of Computing
Scientific Computing and Imaging Institute
University of Utah
WEB 2815
72 Central Campus Drive, Salt Lake City, UT, 84112
Phone (801) 587-3206
Fax (801) 585-6513
Home Page: http://www.sci.utah.edu/~shireen

> On Oct 17, 2019, at 6:38 AM, ccompanionib@gmail.com wrote:
>
> I have been trying to use ShapeWorks tools for some shape analysis
> (semicircular canals of the vestibular system). I am using the python-scripted
> of the ellipsoid example with my data, but after the grooming regardless of
> what parameters I change,
> whether its skipping grooming altogether, filling holes, smoothing,
> antialiasing, etc or any variation thereof,the shape disappear.
> I also try with ShapeWorksStudio changing parameters on the grooming step, and
> it seem to work but then it crashes on the optimization.
> I would really appreciate any help you could give me.


<label_SemicircularCanals.zip>



--
Sincerely,

Claudia Companioni Brito



--
Sincerely,

Claudia Companioni Brito


--
Sincerely,

Claudia Companioni Brito


--
Sincerely,

Claudia Companioni Brito

Attachment: Screenshot from 2020-11-04 18-10-20.png
Description: PNG image



  • [shapeworks-users] DeepSSM, Claudia Companioni-Brito, 11/04/2020

Archive powered by MHonArc 2.6.18.

Top of page