<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=gb2312">
<META content="MSHTML 6.00.6000.16587" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>Hi, Luis,<BR><BR>The dataset I used is "sphere.zip",
which can be downloaded from <A
href="http://appsrv.cse.cuhk.edu.hk/~dfwang/tmp/sphere.zip">http://appsrv.cse.cuhk.edu.hk/~dfwang/tmp/sphere.zip</A>
(the size is about 30K). By uncompressing this file, you will see Analyze file
of "sphere.hdr" and "sphere.img".<BR><BR>The parameters I used are listed as
follows,<BR><BR>Internal Forces (Alpha): 0.8<BR>External Forces (Beta):
0.8<BR>Damping Forces (Gamma): 0.35<BR>Range of Search: 2<BR>Rigidity
(Regularizing): 0.5<BR>Number of Iterations: 1<BR><BR>The parameters involved in
the gradient calculation of <BR>DeformableModelSimplexMesh are keep unchanged.
That is, I did not make any <BR>change to the source code of
DeformableModelSimplexMesh.<BR><BR>The steps to reproduce my results are listed
as follows,<BR><BR>1. "File/Load File", select "sphere.hdr";<BR>2. left click
the center point of the upper left image, then click <BR>"File/Create Mesh", you
will see a simplex spherical mesh in the window;<BR>3. "File/Preprocess Image",
wait about two minutes to finish the preprocess;<BR>4. Click "Settings/Deform
Mesh" about 25 times. Each time after your click, <BR>you will see that the
simplex mesh deform. In this process, you can see the <BR>problem I described in
my last email.<BR><BR>By the way, if I set the parameter of "Number of
Iterations: 1" to 25, and <BR>click "Settings/Deform Mesh" only once. The result
I got seems quite <BR>similiar to the initial mesh, and is not like the one got
by seting "Number <BR>of Iterations: 1" to 1 and click "Settings/Deform Mesh" 25
times. I think <BR>this is another bug in this program.<BR><BR>If any part in
the above desription is not clear enough, please let me know <BR>and I will try
to provide more details.<BR><BR>Best wishes,<BR>Defeng<BR><BR><BR>----- Original
Message ----- <BR>From: "Luis Ibanez" <<A
href="">luis.ibanez@kitware.com</A>><BR>To: "Defeng WANG" <<A
href="">dfwang@cse.cuhk.edu.hk</A>><BR>Cc: "Leila Baghdadi" <<A
href="">baghdadi@phenogenomics.ca</A>>; <BR><<A
href="">insight-developers@itk.org</A>>; <<A
href="">insight-users@itk.org</A>><BR>Sent: Friday, February 01, 2008 10:40
AM<BR>Subject: Re: [Insight-developers] bug report on DeformableModelSimplexMesh
<BR>application<BR><BR><BR>><BR>> Hi Defeng,<BR>><BR>> Can you
please post in a public web site the image that<BR>> you are using as input,
and the full set of parameters<BR>> that you are using for running this
application ?<BR>><BR>> Otherwise it is very hard for us to figure out how
far<BR>> from the edge of the image you are seeing the final<BR>> contour
stopping.<BR>><BR>> Please give use a very clear sequence of step that
we<BR>> could follow to reproduce your experience
here.<BR>><BR>><BR>><BR>>
Thanks<BR>><BR>><BR>>
Luis<BR>><BR>><BR>> -------------------<BR>> Defeng WANG
wrote:<BR>>> Hi, Leila,<BR>>><BR>>> Thanks for your reply. It
seems that the parameters in the UI of<BR>>> DeformableModelSimplexMesh
are not related to the gradient calculation. I<BR>>> read the source code
involved in DeformableModelSimplexMesh. As for the<BR>>> gradient
calculation, three filters are used successively as
follows,<BR>>><BR>>>
m_GradientMagnitude->SetInput(m_CastImage->GetOutput()
);<BR>>>
m_GradientMagnitude->SetSigma(0.5);<BR>>><BR>>>
m_SigmoidImage->SetInput(
m_GradientMagnitude->GetOutput());<BR>>>
m_SigmoidImage->SetOutputMinimum(0);<BR>>>
m_SigmoidImage->SetOutputMaximum(1);<BR>>>
m_SigmoidImage->SetAlpha(230);<BR>>>
m_SigmoidImage->SetBeta(1300);<BR>>><BR>>>
m_GradientFilter->SetInput(
m_SigmoidImage->GetOutput());<BR>>>
m_GradientFilter->SetSigma( 0.5);<BR>>><BR>>><BR>>> I have
taken a look at the final gradent image. It seems that it is good<BR>>>
enough to describe the<BR>>> edge or boundary of a sphere. So what do you
think of this problem <BR>>> existing<BR>>> in this simple
example?<BR>>><BR>>><BR>>> Look forard to your
response.<BR>>><BR>>> Best wishes,<BR>>>
Defeng<BR>>><BR>>> ----- Original Message ----- From: "Leila
Baghdadi"<BR>>> <<A
href="">baghdadi@phenogenomics.ca</A>><BR>>> To: "Defeng WANG" <<A
href="">dfwang@cse.cuhk.edu.hk</A>><BR>>> Cc: <<A
href="">insight-developers@itk.org</A>>; <<A
href="">insight-users@itk.org</A>><BR>>> Sent: Thursday, January 31,
2008 10:54 PM<BR>>> Subject: Re: [Insight-developers] bug report
on<BR>>> DeformableModelSimplexMesh
application<BR>>><BR>>><BR>>>> Hi
Defeng,<BR>>>><BR>>>> Yes I have used that code many a times.
I am not sure if I understand<BR>>>> what you mean by "it does not
converge".<BR>>>><BR>>>> My understanding of deformable models
is<BR>>>><BR>>>> 1. you must make sure your gradient image is
created properly. Use<BR>>>> paraview to look at the vector
image<BR>>>><BR>>>> 2. you must play with the parameters to
get the model to converge,<BR>>>><BR>>>> This is a specific
type of deformable models which uses simplex mesh.<BR>>>> I suggest you
read about this algorithm first. This is developed based<BR>>>> on the
paper by Herve Delingette of INRIA
france.<BR>>>><BR>>>><BR>>>>
Leila<BR>>>><BR>>>> On Thu, 2008-31-01 at 22:29 +0800, Defeng
WANG wrote:<BR>>>><BR>>>>> Hello, ITK users and
developers,<BR>>>>><BR>>>>> I found that
DeformableModelSimplexMesh, provided in<BR>>>>>
InsightApplications-3.4.0, does not converge. It seems that
the<BR>>>>> deformable mesh will not stop near the boundary of one
object to be<BR>>>>> segmented. I tried a very simple volume data
set containing a sphere<BR>>>>> only. The initial mesh I used is a
spherical simplex mesh inside it.<BR>>>>> However, by increasing the
number of iterations, the mesh will<BR>>>>> continue to expand until
it reaches outside of the image and the<BR>>>>> program reports
error.<BR>>>>><BR>>>>> Is there anybody tried this
program successfully before? Or anybody<BR>>>>> knows how to adjust
the parameters properly to get right
segmentation?<BR>>>>><BR>>>>> Best
wishes,<BR>>>>> Defeng<BR>>>>>
_______________________________________________<BR>>>>>
Insight-developers mailing list<BR>>>>> <A
href="">Insight-developers@itk.org</A><BR>>>>> <A
href="">http://www.itk.org/mailman/listinfo/insight-developers</A><BR>>><BR>>><BR>>>
_______________________________________________<BR>>> Insight-developers
mailing list<BR>>> <A href="">Insight-developers@itk.org</A><BR>>>
<A
href="">http://www.itk.org/mailman/listinfo/insight-developers</A><BR>>>
<BR></BODY></HTML>