[ITK-users] Interpolation, Requested Region, and periodic signals

Bradley Lowekamp blowekamp at mail.nih.gov
Thu Apr 14 09:26:29 EDT 2016


Hello,

I think the simplest thing to do would be just to “pad” you time series appropriately. You would just need to append the first image to the back. I think the WrapPadImageFilter should be able to do this.

HTH,
Brad

> On Apr 14, 2016, at 8:09 AM, Cyril Mory <cyril.mory at uclouvain.be> wrote:
> 
> Hi ITK users,
> 
> I am computing 4D images of the breathing thorax. The first 3 dimensions are the spatial ones, the last one is the respiratory phase (breathing is considered periodic). At some point in the algorithm, I have to interpolate a 4D sequence of volumes along the phase dimension to create a 3D volume. I use a custom interpolation filter, not the ITK one, with the GenerateData implemented in CUDA (I work on RTK, a library based on ITK that has support for CUDA).
> 
> I would like to request only the required data for this interpolation, in order to avoid keeping the whole 4D sequence in GPU memory. Most of the time, the two 3D volumes between which I need to interpolate are consecutive in the 4D sequence, so it is easy to write a GenerateInputRequestedRegion() that requests a 4D region of size 2 on the fourth dimension.
> But when interpolating between the last and the first 3D volumes of the sequence (remember : breathing is periodic), the only ITK::ImageRegion that contains them both is the whole 4D sequence. I would like to avoid loading the whole 4D sequence.
> 
> Any suggestions on how I could do that ?
> 
> Best regards,
> Cyril
> _____________________________________
> 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.php
> 
> 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://public.kitware.com/mailman/listinfo/insight-users



More information about the Insight-users mailing list