<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style>
</head>
<body class='hmmessage'>
<br>As a follow-up to my earlier message, I can notice that the MSE & Harmonic energy tapers off after just 20 to 30 iterations. This is pretty fast compared to what I have seen with other methods and implementations. However, I have not checked with all permutation and combination of parameters that are possible here.<br>I suppose these are the only 2 metrics that we need to watch, apart from making sure that the Jacobian stays positive. I am using the same parameters given earlier with Deformation field sigma = 1.5 (default)<br>I can also see that adding to the multi-resolution levels seems to help, mainly in getting a higher harmonic energy. <br><br>Apart from viewing the resulting image and the transformation, I would like to measure the registration error which helps in
understanding how "good" is the warping. One
method is to pick a few fiducial point pairs in the 2 original images and apply the transformation field onto the fiducials of the fixed image. Is there a way to do so here? I checked with the example code and could not find any.<br><br>Appreciate your comments... thanks,<br>Ganesh<br><br>> <br>> > From: tom.vercauteren@m4x.org<br>> > Date: Thu, 24 Feb 2011 08:47:01 +0100<br>> > Subject: Re: [Insight-users] Demons Registration<br>> > To: nganesh76@hotmail.com<br>> > CC: insight-users@itk.org<br>> > <br>> > Hi Ganesh,<br>> > <br>> > On Wed, Feb 23, 2011 at 19:04, Ganesh Narayanasamy<br>> > <nganesh76@hotmail.com> wrote:<br>> > ><br>> > > When I run DemonsRegistration using the options as given in the command line<br>> > > below, I can get a few output files like final image, deformation field etc.<br>> > > How would I get the rest of the output like "True field file"?<br>> > <br>> > The "true field" is an input not an output. It may be used for<br>> > synthetic experiments where the actual deformation is known.<br>> > DemonsRegistration then compares at each iteration its result against<br>> > this ground truth (but of course it does not use it to drive the<br>> > registration).<br>> > <br>> > <br>> > > As the image was not deforming enough, I had opted to reduce sigma from the<br>> > > default value of 1.5 to 0.8.. is this the correct approach?<br>> > <br>> > This entirely depends on your use case... That being said, I would<br>> > first try to use more iterations and then reduce the sigma.<br>> > <br>> > <br>> > > Size of the fixed and moving images are 384x384x64 pixels each. I would like<br>> > > to register using the multi-resolution approach starting with 1/4th number<br>> > > of pixels in all 3 axes (i.e, 96x96x16), going to 1/2 (ie., 192x192x32<br>> > > pixels) and finally to the full image. How will I do this?<br>> > <br>> > It is already done, see the number of multiresolution levels and<br>> > number of iterations per level:<br>> > > Number of multiresolution levels: 3<br>> > > Number of demons iterations: [ 15 10 5 ]<br>> > <br>> > Hope this helps,<br>> > Tom<br>> > <br>> > <br>> > > Command:<br>> > >>DemonsRegistration.exe -f CBCT1.mha -m CBCT4.mha -o Final.mha -O DefFld.mha<br>> > >> -s 0.8 -d 1<br>> > ><br>> > > Starting demons registration with the following arguments:<br>> > > Arguments structure:<br>> > > Fixed image file: CBCT1.mha<br>> > > Moving image file: CBCT4.mha<br>> > > Input field file:<br>> > > Input transform file:<br>> > > Output image file: Final.mha<br>> > > Output field file: DefFld.mha<br>> > > True field file:<br>> > > Number of multiresolution levels: 3<br>> > > Number of demons iterations: [ 15 10 5 ]<br>> > > Deformation field sigma: 0.8<br>> > > Update field sigma: 0<br>> > > Maximum update step length: 2<br>> > > Update rule: exponentialize and compose (Diffeomorphic)<br>> > > Type of gradient: symmetrized (ESM for diffeomorphic and compositive)<br>> > > Use histogram matching: false<br>> > > Algorithm verbosity (debug level): 1<br>> > ><br>> > > Thanks,<br>> > > Ganesh<br>> <br><br> </body>
</html>