[Rtk-users] 3D TV regularization

Cyril Mory cyril.mory at creatis.insa-lyon.fr
Thu Jun 19 05:32:14 EDT 2014


Dear Uffe,

Simply increasing the size of the reconstructed volume until the object 
is fully contained in it should do the trick. An intuitive way to 
understand this is the following: the algorithm tries to "explain" the 
attenuation it sees on the projections. If the data has been partially 
truncated, some rays have "extra attenuation" that has to go somewhere 
in the reconstructed volume. The center of the reconstructed volume is 
constrained by all the other rays, so this "extra attenuation" is pushed 
to the borders. If, on the other hand, the reconstructed volume is large 
enough, this "extra attenuation" can go where it belongs, and does not 
cause hyperattenuation on the borders.

Note that you can still apply a crop to FOV at the very end of the 
reconstruction process.

I hope it helps
Regards,
Cyril

On 06/19/2014 11:02 AM, Uffe Bernchou wrote:
>
> Dear Cyril,
>
> We are currently playing with the 3D TV regularization iterative 
> reconstruction implemented in RTK for our cone beam data. The results 
> look promising indeed.
>
> One thing that worries me a little is a ring artefact we observe in 
> the reconstructions (please see attached image). This ring artefact we 
> usually remove in the standard fdk reconstruction by using padding. 
> Does the TV regularization iterative reconstruction have a similar 
> feature for removing the ring artefact?
>
> Best regards
>
> Uffe Bernchou
>
> ------------------------------------------------------------------------
>
> *Fra:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr]
> *Sendt:* 3. juni 2014 17:30
> *Til:* Uffe Bernchou; 'Simon Rit'
> *Cc:* rtk-users at openrtk.org
> *Emne:* Re: [Rtk-users] Release of RTK v1.0.0
>
> Dear Uffe,
>
> The method is described and compared to others in this PhD thesis :
> http://tel.archives-ouvertes.fr/docs/00/98/57/28/PDF/TH2014_Mory_Cyril.pdf
> That's for the pdf. You can access the description here :
> http://tel.archives-ouvertes.fr/tel-00985728
>
> Your comments/questions are welcome.
>
> Best regards,
> Cyril
>
> Le 03/06/2014 14:36, Uffe Bernchou a écrit :
>
>> Dear Simon,
>>
>> Do you have a reference publication for the new iterative 3D 
>> reconstruction with 3D TV regularization implemented in the v.1.00 
>> release of RTK?
>>
>> Best regards,
>>
>> Uffe
>>
>> ------------------------------
>>
>> Uffe Bernchou, Ph.D.
>>
>> Postdoc
>>
>> Laboratory of Radiation Physics, Odense University Hospital
>>
>> Institute of Clinical Research, University of Southern Denmark
>>
>> Email: uffe.bernchou at rsyd.dk <mailto:uffe.bernchou at rsyd.dk>
>>
>> Phone: +45 29796478
>>
>> Sdr. Boulevard 29, Indgang 97, 5000 Odense C
>>
>> ------------------------------------------------------------------------
>>
>> *Fra:*Rtk-users [mailto:rtk-users-bounces at openrtk.org] *På vegne af 
>> *Simon Rit
>> *Sendt:* 28. april 2014 16:33
>> *Til:* rtk-users at openrtk.org <mailto:rtk-users at openrtk.org>
>> *Emne:* [Rtk-users] Release of RTK v1.0.0
>>
>> Dear RTK users,
>>
>> It is my pleasure to announce the release of RTK v1.0.0. As announced 
>> previously on the mailing list, the new release is no more compatible 
>> with ITK 3.20, this release is compatible with ITK v4.2.0 and more 
>> recent releases. We are still actively developing RTK with both new 
>> filtered backprojection and iterative algorithms so I still encourage 
>> the use of git to be up-to-date. As usual, please report any issue 
>> that you encounter with RTK, we are happy to help when we can. Note 
>> the list of publications using RTK 
>> <http://www.openrtk.org/RTK/project/publications.php> on the main 
>> site: please send me the DOI of your articles if you want them to 
>> appear there. We have also added a webpage with use cases 
>> <http://www.openrtk.org/RTK/project/casestudies.html>.
>>
>> I'm looking forward to your feedback. Regards,
>>
>> Simon
>>
>>
>> Release notes:
>>
>> - introduction of the itk::CudaImage object that better handles the 
>> memory transfers between the CPU and the GPU
>>
>> - pre-processing for projection images: binning, median filter, 
>> lookup-table filter
>>
>> - filtered backprojection with offset center-of-rotation based on 
>> [Gullberg et al, TMI, 1986]
>>
>> - improved Joseph forward projector and implementation of its exact 
>> adjoint
>>
>> - improved SART reconstruction [Andersen, Ultrason Imaging, 1984]
>> - new iterative 3D reconstruction using conjugate gradient optimization
>>
>> - new iterative 3D reconstruction with 3D TV regularization using the 
>> Alternating Direction Method of Multipliers combined with the 
>> Augmented Lagrangian method
>>
>> Documentation:
>>
>> - improved Doxygen with dot graphs to describe mini-pipelines, e.g., 
>> in the SART page 
>> <http://www.openrtk.org/Doxygen/classrtk_1_1SARTConeBeamReconstructionFilter.html>
>>
>> - updated consortium 
>> <http://www.openrtk.org/RTK/project/consortium.html> and alternative 
>> platforms <http://www.openrtk.org/RTK/resources/alternatives.html> pages
>>
>> - updated wiki
>>
>>
>>
>>
>> _______________________________________________
>> Rtk-users mailing list
>> Rtk-users at openrtk.org  <mailto:Rtk-users at openrtk.org>
>> http://public.kitware.com/cgi-bin/mailman/listinfo/rtk-users
>
>
>
> -- 
> Cyril Mory, Post-doc
> CREATIS
> Leon Berard cancer treatment center
> 28 rue Laënnec
> 69373 Lyon cedex 08 FRANCE
>   
> Mobile: +33 6 69 46 73 79
>
> ------------------------------------------------------------------------
>
> <http://www.avast.com/>
>
> 	
>
> Ce courrier électronique ne contient aucun virus ou logiciel 
> malveillant parce que la protection Antivirus avast! 
> <http://www.avast.com/> est active.
>

-- 
--
Cyril Mory, Post-doc
CREATIS
Leon Berard cancer treatment center
28 rue Laënnec
69373 Lyon cedex 08 FRANCE

Mobile: +33 6 69 46 73 79

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20140619/3aa3149d/attachment-0009.html>


More information about the Rtk-users mailing list