[Elastix] FW: Trouble with large BSpline registrations

S. Klein s.klein at erasmusmc.nl
Tue Apr 7 13:22:12 CEST 2015


Hi Robin,
Good to hear that elastix was useful for you.

To solve your problem, you could try to set the NumberOfJacobianMeasurements in the parameter file:

NumberOfJacobianMeasurements: The number of voxels M where the Jacobian is measured, which is used to estimate the covariance matrix. The parameter can be specified for each resolution, or for all resolutions at once.
example: (NumberOfJacobianMeasurements 5000 10000 20000)
Default value: M = max( 1000, nrofparams ), with nrofparams the number of transform parameters. This is a rather crude rule of thumb, which seems to work in practice. In principle, the more the better, but the slower. The parameter has only influence when AutomaticParameterEstimation is used.

But do you really (intend to) use more than 2billion transform parameters? what is your control point spacing/size?

Cheers,
Stefan

On 07-04-2015 02:02, Robin.Pearce at csiro.au<mailto:Robin.Pearce at csiro.au> wrote:
Hello elastix team,

I would like to start by saying thank you for an excellent software package which has made my life so much easier. Unfortunately I’ll be moving to a different field shortly and probably won’t be working with images anymore, but I’ve very much enjoyed using elastix.

I just wanted to let you know of a problem that I’ve had while using the BSpline registration. I’ve probably been pushing the limits on what this software can handle, as I’ve been using 3D images about 5GB in size, but I’ve noticed that the software crashes whenever the parameter “NumberOfJacobianMeasurements” is larger than 2^31. This seems to me like an indexing problem, something I’ve encountered in other large computational problems before. First of all, is there a quick fix to this? And by quick, I mean in the next two weeks, because after that I won’t have enough time to use it for what I want. Otherwise, I’m just leaving this here as a comment for potential further development.

Thanks again for the excellent software,

Robin.



_______________________________________________
Elastix mailing list
Elastix at bigr.nl<mailto:Elastix at bigr.nl>
http://lists.bigr.nl/cgi-bin/mailman/listinfo/elastix



--
Stefan Klein
+31 10 7043442
http://www.bigr.nl/people/StefanKlein
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bigr.nl/pipermail/elastix/attachments/20150407/671a2cb9/attachment.html>


More information about the Elastix mailing list