[Rtk-users] Angular gaps

[Simon Rit]

Hi Andy,
Unfortunately I'm not going to the AAPM so we'll have to continue
communicating by email. Let us know if there are developments that would be
of interest for RTK.
Regarding your questions:
1. you can not have both a short scan and a displaced detector, otherwise
you have missing data. The consequence of missing data is probably what you
observe here. We could add a check to inform the user but there is
essentially not enough information to reconstruct this.
2. We do weight the projections by uneven angles in the FDK algorithms.
This is done during the 2D weighting. Follow the variable
m_AngularWeightsAndRampFactor in rtkFDKWeightProjectionFilter.txx:
https://github.com/SimonRit/RTK/blob/master/code/rtkFDKWeightProjectionFilter.txx
I hope this helps. Cheers,
Simon


On Wed, May 15, 2013 at 8:15 AM, Andy Shieh <hsieandy at gmail.com> wrote:

> Hi Simon,
>
> First of all, I regret not being able to going to ICCR, otherwise I could
> really thank you for your work in RTK. It has really helped me a lot in my
> research! Fortunately a few of my colleagues got a chance to talk to you
> (Ricky and John, if you remember). Hopefully I'll see you this year in
> AAPM!!
>
> So today I want to ask you about two things.
>
> 1. Failed reconstruction (due to maximum angular gap too big?)
> I noticed that occasionally RTK gives reconstruction that looks like this.
> [image: 內置圖片 1]
> In case the picture is not sent through, the image has a big angular
> region where the intensity is essentially zero when viewed axially. This
> happens when I use displacement binning, and using half-fan data.
> I summarized all my reconstructions and realize this only happens when the
> "maximum angular gap" of the projections in that bin is greater than 20
> degrees. Therefore I assume it's because the half-fan projection on one
> side not being able to find its corresponding projection image on the other
> side. I just want to check with you if that's what's happening here.
>
> 2. There is a paper on direct weighting of uneven angular spaced
> projections.
> http://link.springer.com/chapter/10.1007%2F978-3-642-33415-3_23
> I think the idea is basically lower the weight of closely spaced
> projections.
> Is this already implemented in RTK?
>
> Thanks Simon !! :)
>
> Cheers,
> Andy
>
> _______________________________________________
> Rtk-users mailing list
> Rtk-users at openrtk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/rtk-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20130515/5dc15f10/attachment-0009.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 53548 bytes
Desc: not available
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20130515/5dc15f10/attachment-0009.png>