No subject
Mon Dec 15 09:51:36 EST 2008
unregister residuals than the Bspline registration.
If you are interesting in further pursuing the use of MutualInformation
from the example
DeformableRegistration8.cxx
you may want to try:
A) Increasing the Grid resolution
You can do this by changing:
unsigned int numberOfGridNodesInOneDimension = 5;
since with the current setting you only get a grid of 8x8x8.
This restrict the capacity of the grid for accommodating local
deformations.
By increasing the number, you will have a finer BSpline grid
that should be able to adapt to smaller and more local
deformations.
B) Running the optimizer for more iterations.
Which may imply in this case to do the following:
optimizer->SetCostFunctionConvergenceFactor( 1.e7 ); << make it smaller
optimizer->SetProjectedGradientTolerance( 1e-6 ); << make it smaller
optimizer->SetMaximumNumberOfIterations( 200 ); << make it larger
optimizer->SetMaximumNumberOfEvaluations( 30 ); << make it larger
optimizer->SetMaximumNumberOfCorrections( 5 ); << make it larger
Another way to approach this, is for you to give it a try
at the example
DeformableRegistration15.cxx
which uses a multi-resolution approach to this same type of
BSpline deformable registration.
This example will use several levels of BSpline grids, in such a
way that you can model small local deformations, but without
having to compute them from the very beginning.
Please give it a try at this latter example,
and let us know what you find,
Thanks
Luis
---------------------------------------------------------------------
On Wed, Feb 11, 2009 at 10:44 AM, marco giordano <marco.giord at gmail.com>wrote:
> Hi Luis,
>
> I first want to thank you for the response.
>
> Second I want to apologize because I did not express myself clearly.
>
> I believe that you are right saying that there is not the "Best"
> method, but what I meant by best
> is of course the best that I can get to run "my application".
>
> Anyway I wanted to stress the fact that I am not an expert in
> registration and that to me it is only needed as a preprocessing
> step, thus I would like to spend little time but have a decent result
> that can allow my application to run correctly, but this of course it
> will be hardly possible and I think someone needs to understand how
> things work.
>
> My images are CT scan of the calf.
>
> I hereby attach 3 images of the axial view (WIN=100HU,LEV=0HU), I
> subtracted the 29th timeframe from the 1st timeframe to show the
> motion state. What you see is a contrast enhancement in the left leg
> (vessels and muscles) and noise of course.
>
> -original contains the original DS
> -bspline contains the one after registration with
> DeformableRegistration8.cxx
> -demons contains the one after the registration with
> DeformableRegistration2.cxx
>
> I am quite satisftited with the bspline using the Mutual Information
> but I am still wondering if one can get something with less
> contours by changing some parameters.
>
> Any further help of course is welcome.
>
> Marco G.
>
>
>
>
>
>
>
>
>
>
>
>
>
> 2009/2/11 Luis Ibanez <luis.ibanez at kitware.com>:
> >
> > Hi Marco,
> >
> >
> > Thanks for the detailed description of the problem
> > that you are working on.
> >
> >
> > Please note that:
> >
> >
> > 0) The is *no such thing* as the "most suitable" or the "best"
> > metric for a given registration problem.
> >
> > The popularization of the misguided notion that there is
> > a "best method" for solving image processing problems is
> > the result of the *decadence* of the publishing system in
> > our field. Those who claim that they have a "best method"
> > for any kind of problem are confused or are trying to
> > confuse you; probably to try to sell you something that
> > you don't need to buy.
> >
> > The desire to claim that a method is better than any other
> > one was born from absence of scientific rigor in publications
> > that disregard the importance of reproducibility in the scientific
> > method, and are simply dedicated to fulfilling the needs of
> > academics for filling up their yearly quotas of intellectual
> > "production" by publishing a require N number of papers a year.
> >
> > First of all, *all* methods have parameters, and by changing
> > the numerical values of these parameters you get *entirely*
> > different results. Therefore any claim that "Method A" is
> > better than "Method B", but doesn't list the set of parameters
> > used on each method, has to produced in a context of ignorance,
> > incompetence or malice.
> >
> > An honest an educated claim will instead report
> > something like:
> >
> > Method A (as implemented in the source code that you can
> > download from website Xa), when run on the images K, L
> > (that you can download from the website Y), using the
> > set of parameter (p1=0.5,p2=2.3.....etc) produce the
> > following results (R1), while
> >
> > Method B (as implemented in the source code that you can
> > download from website Xb), when run on the images K, L
> > (that you can download from the website Y), using the
> > set of parameter (q1=0.7,q2=7.3.....etc) produce the
> > following results (Z1)
> >
> > and in the context of the application (AA: which can be
> > radiation treatment, or surgical planning, or teaching,
> > ...) we consider that the result R1 is better than Z1.
> >
> > That is a serious technical report,
> > while the simple claim:
> >
> > "Method A" is better than "Method B"
> >
> > is simply a *superstition*.
> >
> >
> > Please,
> > demand from authors and editor of technical publications
> > to raise their standards to the at least the basic
> > scientific level that undergrad students will learn in
> > Physics 101.
> >
> > Please,
> > Refuse to collaborate in the dissemination of superstitious
> > beliefs.
> >
> >
> >
> > That being said,
> >
> >
> > 1) Mean squares may still be able to register your images,
> > if the propagation of the contrast agent between Image N
> > and image N-1 is not too large, and if there are enough
> > regions of high intensity contrast (not agent contrast)
> > that could guide the metric down the registration process.
> >
> > But of course, only a set of experiments that include
> > a process of parameter fine-tunning will tell for sure
> > if that is the case for your specific set of images.
> >
> >
> > 2) The Demons algorithm is essentially using a Mean Squares
> > metric. The motion that it computes is represented as
> > a vector field, where you get a displacement vector per
> > pixel, and the vector field is smoothed using Gaussians.
> >
> >
> > 3) BSpline and MutualInformation: Both of these classes
> > have a set of about ten parameters combined
> >
> > * number of grid point in the Spline
> > * spline order
> > * number of samples in the metric
> > * number of histogram bins
> > * Sigmas for the distributions
> >
> > not to mention the parameters of the optimizer that
> > are typically 3 to 5.
> >
> > If the registration is not capturing large deformations,
> > you may have to consider a multi-resolution approach to
> > the BSpline grid, such as the one used in
> >
> > DeformableRegistration15.cxx
> >
> >
> > 4) You may also find useful to try the Demons multi-resolution
> > examples in
> >
> > DeformableRegistration16.cxx
> > DeformableRegistration17.cxx
> >
> >
> >
> > Specific suggestions will require that you tell us more
> > about the images that you are registering
> >
> > Anatomy : (brain ? lung ? liver ?)
> > Modality : MRI ? CT ?
> >
> > Screen shots of the images and the current results would
> > be great.
> >
> > BTW, you may be interested in trying the application VV
> > http://www.midasjournal.org/browse/publication/288
> > that is designed for visualizing the result of these kind
> > of registration problems.
> >
> >
> >
> > Regards,
> >
> >
> >
> > Luis
> >
> >
> >
> > ---------------------
> > marco giordano wrote:
> >>
> >> Hi all,
> >>
> >> I am doing a registration of a 3D temporal sequence that consist of 40
> >> frames (Vol1 .... Vol40).
> >> .
> >> The sequence shows the contrast uptake in vessels and human tissues
> >> and during the sequence some squeezing and non rigid motion occurs .
> >>
> >> Normally I register all the frame ( Vol2 ... Vol 40) to the first
> >> frame (Vol1) or use a concatenation of registration (register VolN to
> >> Vol(N-1)registered)
> >>
> >> For such kind of problem I heard that the Mutual Information is the
> >> most suitable metric.
> >>
> >> In fact a metric as meansquaredifference would try to minimize the
> >> difference in intensities, but the intensities are already different
> >> for
> >> the fact that contrast changes the intensity in each frame.
> >>
> >> So far I tried:
> >>
> >> -DemonsAlgorithm DeformableRegistration2.cxx (registration is good but
> >> the some changes occur in the intensities )
> >> Anyway I do not undestand what kind of metric is used here and for
> >> what kind of motion is most suitable for this algorithm
> >>
> >> -Bspline with Mutual Information DeformableRegistration8.cxx ( here
> >> strong motion is not completely removed ).
> >>
> >> I wanted to ask if anyone has suggestion on how to carry on and if
> >> there is some parameters that can be changed
> >> e.g:
> >>
> >> -the grid spaces etc.
> >> -number of iteration
> >> -the optimizer
> >>
> >> Any help would be appreciated
> >>
> >> PS: I would like to cite ITK whenever the results are suitable for an
> >> article.
> >>
> >> Thanks
> >>
> >
>
>
>
> --
> Marco Giordano
> MSN:gilmour812002 at yahoo.it <MSN%3Agilmour812002 at yahoo.it>
> ICQ :285-118-610
> SKYPE:marcogiord81
>
--0015175ce13249b9b20462fd679c
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
<br>Hi Marco,<br><br>Thanks a lot for posting your images,<br>they are quit=
e illustrative.<br><br>From your images, it looks like the Demons algorithm=
is leaving less <br>unregister residuals than the Bspline registration.<br=
>
<br>If you are interesting in further pursuing the use of MutualInformation=
<br>from the example<br><br>  =
; DeformableRegistration8.cxx<br><br>you =
may want to try:<br><br><br>A) Increasing the Grid resolution<br>
<br> You can do this by changing:<br><br> un=
signed int numberOfGridNodesInOneDimension =3D 5;<br><br> =
since with the current setting you only get a grid of 8x8x8.<br><br> =
This restrict the capacity of the grid for accommodating local=
<br>
deformations.<br><br> By increasing the num=
ber, you will have a finer BSpline grid<br> that should b=
e able to adapt to smaller and more local <br> deformatio=
ns.<br><br><br>B) Running the optimizer for more iterations.<br>
Which may imply in this case to do the following:<br><b=
r> optimizer->SetCostFunctionConvergenceFactor( 1.e7 ); <=
< make it smaller<br> optimizer->SetProjectedGradientTolerance(=
1e-6 ); << make it smaller<br>
optimizer->SetMaximumNumberOfIterations( 200 ); << make it =
larger<br> optimizer->SetMaximumNumberOfEvaluations( 30 ); &=
lt;< make it larger<br> optimizer->SetMaximumNumberOfCorrection=
s( 5 ); << make it larger <br>
<br><br>Another way to approach this, is for you to give it a try<br>at the=
example<br><br> DeformableReg=
istration15.cxx<br><br>which uses a multi-resolution approach to this same =
type of<br>BSpline deformable registration.<br>
<br>This example will use several levels of BSpline grids, in such a<br>way=
that you can model small local deformations, but without<br>having t=
o compute them from the very beginning.<br><br><br>Please give it a try at =
this latter example,<br>
and let us know what you find,<br><br><br> &nb=
sp; Thanks<br><br><br>  =
; Luis<br><br><br><br>-----------------------------------=
----------------------------------<br><div class=3D"gmail_quote">On Wed, Fe=
b 11, 2009 at 10:44 AM, marco giordano <span dir=3D"ltr"><<a href=3D"mai=
lto:marco.giord at gmail.com">marco.giord at gmail.com</a>></span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"border-left: 1px solid rgb(204, =
204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi Luis,<br>
<br>
I first want to thank you for the response.<br>
<br>
Second I want to apologize because I did not express myself clearly.<br>
<br>
I believe that you are right saying that there is not the "Best"<=
br>
method, but what I meant by best<br>
is of course the best that I can get to run "my application".<br>
<br>
Anyway I wanted to stress the fact that I am not an expert in<br>
registration and that to me it is only needed as a preprocessing<br>
step, thus I would like to spend little time but have a decent result=
<br>
that can allow my application to run correctly, but this of course it<br>
will be hardly possible and I think someone needs to understand how<br>
things work.<br>
<br>
My images are CT scan of the calf.<br>
<br>
I hereby attach 3 images of the axial view (WIN=3D100HU,LEV=3D0HU), I<br>
subtracted the 29th timeframe from the 1st timeframe to show the<br>
motion state. What you see is a contrast enhancement in the left leg<br>
(vessels and muscles) and noise of course.<br>
<br>
-original contains the original DS<br>
-bspline contains the one after registration with DeformableRegistration8.c=
xx<br>
-demons contains the one after the registration with<br>
DeformableRegistration2.cxx<br>
<br>
I am quite satisftited with the bspline using the Mutual Information<br>
but I am still wondering if one can get something with less<br>
contours by changing some parameters.<br>
<br>
Any further help of course is welcome.<br>
<br>
Marco G.<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
2009/2/11 Luis Ibanez <<a href=3D"mailto:luis.ibanez at kitware.com">luis.i=
banez at kitware.com</a>>:<br>
<div><div></div><div class=3D"Wj3C7c">><br>
> Hi Marco,<br>
><br>
><br>
> Thanks for the detailed description of the problem<br>
> that you are working on.<br>
><br>
><br>
> Please note that:<br>
><br>
><br>
> 0) The is *no such thing* as the "most suitable" or the &quo=
t;best"<br>
> metric for a given registration problem.<br>
><br>
> The popularization of the misguided notion that there is<br>
> a "best method" for solving image processing problems=
is<br>
> the result of the *decadence* of the publishing system in<br>
> our field. Those who claim that they have a "best me=
thod"<br>
> for any kind of problem are confused or are trying to<br>
> confuse you; probably to try to sell you something that<br>
> you don't need to buy.<br>
><br>
> The desire to claim that a method is better than any other<br>
> one was born from absence of scientific rigor in publications<b=
r>
> that disregard the importance of reproducibility in the scienti=
fic<br>
> method, and are simply dedicated to fulfilling the needs of<br>
> academics for filling up their yearly quotas of intellectual<br=
>
> "production" by publishing a require N number of pape=
rs a year.<br>
><br>
> First of all, *all* methods have parameters, and by changing<br=
>
> the numerical values of these parameters you get *entirely*<br>
> different results. Therefore any claim that "Method A"=
; is<br>
> better than "Method B", but doesn't list the set =
of parameters<br>
> used on each method, has to produced in a context of ignorance,=
<br>
> incompetence or malice.<br>
><br>
> An honest an educated claim will instead report<br>
> something like:<br>
><br>
> Method A (as implemented in the source code that you can=
<br>
> download from website Xa), when run on the images K, L<b=
r>
> (that you can download from the website Y), using the<br=
>
> set of parameter (p1=3D0.5,p2=3D2.3.....etc) produce the=
<br>
> following results (R1), while<br>
><br>
> Method B (as implemented in the source code that you can=
<br>
> download from website Xb), when run on the images K, L<b=
r>
> (that you can download from the website Y), using the<br=
>
> set of parameter (q1=3D0.7,q2=3D7.3.....etc) produce the=
<br>
> following results (Z1)<br>
><br>
> and in the context of the application (AA: which can be<=
br>
> radiation treatment, or surgical planning, or teaching,<=
br>
> ...) we consider that the result R1 is better than Z1.<b=
r>
><br>
> That is a serious technical report,<br>
> while the simple claim:<br>
><br>
> "Method A" is bette=
r than "Method B"<br>
><br>
> is simply a *superstition*.<br>
><br>
><br>
> Please,<br>
> demand from authors and editor of technical publications<br>
> to raise their standards to the at least the basic<br>
> scientific level that undergrad students will learn in<br>
> Physics 101.<br>
><br>
> Please,<br>
> Refuse to collaborate in the dissemination of superstitious<br>
> beliefs.<br>
><br>
><br>
><br>
> That being said,<br>
><br>
><br>
> 1) Mean squares may still be able to register your images,<br>
> if the propagation of the contrast agent between Image N<br>
> and image N-1 is not too large, and if there are enough<br>
> regions of high intensity contrast (not agent contrast)<br>
> that could guide the metric down the registration process.<br>
><br>
> But of course, only a set of experiments that include<br>
> a process of parameter fine-tunning will tell for sure<br>
> if that is the case for your specific set of images.<br>
><br>
><br>
> 2) The Demons algorithm is essentially using a Mean Squares<br>
> metric. The motion that it computes is represented as<br>
> a vector field, where you get a displacement vector per<br>
> pixel, and the vector field is smoothed using Gaussians.<br>
><br>
><br>
> 3) BSpline and MutualInformation: Both of these classes<br>
> have a set of about ten parameters combined<br>
><br>
> * number of grid point in the Spline<br>
> * spline order<br>
> * number of samples in the metric<br>
> * number of histogram bins<br>
> * Sigmas for the distributions<br>
><br>
> not to mention the parameters of the optimizer that<br>
> are typically 3 to 5.<br>
><br>
> If the registration is not capturing large deformations,<br>
> you may have to consider a multi-resolution approach to<br>
> the BSpline grid, such as the one used in<br>
><br>
> DeformableRegistration15.cxx<br>
><br>
><br>
> 4) You may also find useful to try the Demons multi-resolution<br>
> examples in<br>
><br>
> DeformableRegistration16.cxx=
<br>
> DeformableRegistration17.cxx=
<br>
><br>
><br>
><br>
> Specific suggestions will require that you tell us more<br>
> about the images that you are registering<br>
><br>
> Anatomy : (brain ? lung ? liver ?)<br>
> Modality : MRI ? CT ?<br>
><br>
> Screen shots of the images and the current results would<br>
> be great.<br>
><br>
> BTW, you may be interested in trying the application VV<br>
> <a href=3D"http://www.midasjournal.org/browse/publication/288" target=
=3D"_blank">http://www.midasjournal.org/browse/publication/288</a><br>
> that is designed for visualizing the result of these kind<br>
> of registration problems.<br>
><br>
><br>
><br>
> Regards,<br>
><br>
><br>
><br>
> Luis<br>
><br>
><br>
><br>
> ---------------------<br>
> marco giordano wrote:<br>
>><br>
>> Hi all,<br>
>><br>
>> I am doing a registration of a 3D temporal sequence that consist o=
f 40<br>
>> frames (Vol1 .... Vol40).<br>
>> .<br>
>> The sequence shows the contrast uptake in vessels and human tissue=
s<br>
>> and during the sequence some squeezing and non rigid motion occurs=
.<br>
>><br>
>> Normally I register all the frame ( Vol2 ... Vol 40) to the first<=
br>
>> frame (Vol1) or use a concatenation of registration (register VolN=
to<br>
>> Vol(N-1)registered)<br>
>><br>
>> For such kind of problem I heard that the Mutual Information is th=
e<br>
>> most suitable metric.<br>
>><br>
>> In fact a metric as meansquaredifference would try to minimize the=
<br>
>> difference in intensities, but the intensities are already differe=
nt<br>
>> for<br>
>> the fact that contrast changes the intensity in each frame.<br>
>><br>
>> So far I tried:<br>
>><br>
>> -DemonsAlgorithm DeformableRegistration2.cxx (registration is good=
but<br>
>> the some changes occur in the intensities )<br>
>> Anyway I do not undestand what kind of metric is used here and for=
<br>
>> what kind of motion is most suitable for this algorithm<br>
>><br>
>> -Bspline with Mutual Information DeformableRegistration8.cxx ( her=
e<br>
>> strong motion is not completely removed ).<br>
>><br>
>> I wanted to ask if anyone has suggestion on how to carry on and if=
<br>
>> there is some parameters that can be changed<br>
>> e.g:<br>
>><br>
>> -the grid spaces etc.<br>
>> -number of iteration<br>
>> -the optimizer<br>
>><br>
>> Any help would be appreciated<br>
>><br>
>> PS: I would like to cite ITK whenever the results are suitable for=
an<br>
>> article.<br>
>><br>
>> Thanks<br>
>><br>
><br>
<br>
<br>
<br>
</div></div><font color=3D"#888888">--<br>
Marco Giordano<br>
<a href=3D"mailto:MSN%3Agilmour812002 at yahoo.it">MSN:gilmour812002 at yahoo.it<=
/a><br>
ICQ :285-118-610<br>
SKYPE:marcogiord81<br>
</font></blockquote></div><br>
--0015175ce13249b9b20462fd679c--
More information about the Insight-users
mailing list