[Insight-users] bug in DeformableModelSimplexMesh ?

Tue Jul 27 08:37:04 EDT 2010

Dear Dzenan

As you proposed in your mail, we created an unnamed block and created a new deform filter from scratch 

We instantiate the filter and set all parameters

However, this modification results in the same error message as described below

So our problem is not yet solved

Thank you for your time, kind regards

Jan Menssen 

From: Dženan Zukić [mailto:dzenanz at gmail.com] 
Sent: donderdag 22 juli 2010 14:56
To: Menssen, Jan
Cc: insight-users at itk.org; Nillesen, Maartje
Subject: Re: [Insight-users] bug in DeformableModelSimplexMesh ?

I discovered few weeks back, that you cannot just change input mesh for deform filter, that is

m_DeformFilter->SetInput(m_SimplexMesh);

this line is a problem. The workaround I used was to create unnamed block and create a new deform filter from scratch (instantiate filter and set a bunch of parameters) instead of just setting setting input mesh and calling update.

Regards,

Dženan

On Wed, Jul 21, 2010 at 11:37, <J.Menssen at cukz.umcn.nl> wrote:

Dear All

We are working on a segmentation algorithm using deformable models with a refine mesh. The deformation and refinement is done  in a for loop. Our application is based on the DeformableModel application as given in the InsightApplications example DeformableModelSimplexMesh.

We modified the for loop in de DeformMesh method (DeformableModelApplication.cxx) with a refinement of the mesh, but this crashes. Below a part of (pseudo) code is given.. 

     for i=1:i<nr_iterations;i++

   {

      m_simplexMesh->DisconnectPipeline()

      m_deformFilter-SetInput(m_simplexMesh);

      m_deformFilter->Update(0;

      simplexMesh2 = m_deformFilter->GetOutput();

      m_refineMesh->SetInput(simplexMesh2);

      m_refineMesh->Update();

      m_simplexMesh=m_refineMesh->GetOutput();

  }

m_deformFilter is a DeformableSimplexMesh3DGradientConstraintForceFilter, 

m_RefineMesh is a  itkSimplexMeshAdaptTopologyFilter

To isolate the bug, we started with a little modification of the DeformMesh method in DeformableModelApplication.cxx. The modifications are given below (line 522 in original file, InsightApplciations 3.16). Running this code results also in a crash

*----------------------------------------------------------------------------------------------------------------------------------------

SimplexMeshType::Pointer simplexMesh2 = m_SimplexMesh;

const unsigned int numberOfIterationsToGo = atoi(m_IterationsValueInput->value()); 

for( unsigned int i=0; i<numberOfIterationsToGo; i++ )

{

      std::cout << " Iteration   " << i << std::endl;

//* -------------- start modified code -------------------------------------------

        // THIS IS MODIFIED CODE THAT (WE THINK) DOESN'T VIOLATE THE ITK RULES BUT THIS

 // CODE CRASHES

                        m_SimplexMesh->DisconnectPipeline();

                        m_DeformFilter->SetInput(m_SimplexMesh);

                        m_DeformFilter->SetIterations(1);

                        m_DeformFilter->Update();

                        simplexMesh2 = m_DeformFilter->GetOutput();

                        SimplexMesh2->DisconnectPipeline(0

m_DeformFilter->SetInput(simplexMesh2);

                        m_DeformFilter->SetIterations(1);

                        m_DeformFilter->Update();

                        m_SimplexMesh = m_DeformFilter->GetOutput();

// --------------------- end modified code -------------------------------------- */

      m_SimplexMeshToShow  = m_SimplexMesh;

      this->RefreshMeshVisualization();

      // force a redraw

      axialView->redraw();

      coronalView->redraw();

      sagittalView->redraw();

      surfaceView->redraw();

      Fl::check(); 

 }

*-------------------------------------------------------------------------------------------------------------------------------------------

Error message (using Visual Studio 2005 and/or 2008) of the crash

Debug Assertion Failed !

        Expression map/set iterator not dereferencable 

Do we something wrong?  Using a for loop in a pipeline is dangerous but in bug 0001080, we read it’s possible to use a for loop. Also at the Old Nabble forum  I read a message that is must be possible to use a for loop. (http://old.nabble.com/Using-image-iterators-combined-with-filters-in-a-loop-td28684443.html <http://old.nabble.com/Using-image-iterators-combined-with-filters-in-a-loop-td28684443.html> )

Even if we changed the code in the original example DeformableModelApplication according to the above  referenced Old Nabble article the application crashes (see below)

m_DeformFilter->SetInput( m_SimplexMesh );

      m_DeformFilter->SetIterations(1); 

      m_DeformFilter->Update();

      m_SimplexMesh =  m_DeformFilter->GetOutput(); 

        m_SimplexMesh->DisconnectPipeline();

So is there a bug in the original ITK sources? As far as we now, ITK doesn’t allow to change the input of a filter and we think this is what happened in the original DeformableSimplexMesh application.

Can someone help us with our problem ?

We tested with ITK 2.8, ITK 3.16 and 3.18 and all crashes

 Jan Menssen
Research assistant

Radboud University Nijmegen Medical Centre
Clinical Physics Lab - Peadiatrics
Huispost 833, route 833
P.O  box  9101
6500 HB Nijmegen
Telefoon: +31 24 36 19063
E-mail J.Menssen at cukz.umcn.nl <mailto:J.Menssen at cukz.umcn.nl>  

http://www.umcn.nl <http://www.umcn.nl/> 

Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.

_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users

Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20100727/5c1f8778/attachment-0001.htm>