[Rtk-users] sart for planar tomosynthesis

[Simon Rit]

Hi Robert,
You're taking the problem from the wrong point of view. You should place
the source and the detector in the coordinate system of the object, so we
should not see your object move in your diagram. Move the source and
detector only.
About the size, either the detector is too small or the object is too large
but for sure, in practice and in RTK, both can't change during the
acquisition!
Good luck,
Simon

On Wed, Mar 4, 2015 at 1:55 PM, "Robert Calließ" <Robert.Calliess at gmx.de>
wrote:

>
>
> Hello Simon,
> I try to handle a planar tomosynthesis where the flat panel detectors are
> arranged in a 3x3 matrix. For each view,
> the object and the detector are aligned, so that the "line" from xray
> source through the object center and through the middle
> of the detector. As you told me, I do not change the size of the object.
> The problem is, that the object in the orhogonal view (center)
> covers the whole detector. For any other view some parts of the object are
> outside of the cone beam and so are not part of the
> projection image (see attached pdf). The detector's size is fixed and the
> object's size as well.
>
> Best regards,
> Robert
>
>  *Gesendet:* Dienstag, 03. März 2015 um 16:37 Uhr
> *Von:* "Simon Rit" <simon.rit at creatis.insa-lyon.fr>
> *An:* "Robert Calließ" <Robert.Calliess at gmx.de>
> *Cc:* "rtk-users at openrtk.org" <rtk-users at openrtk.org>
> *Betreff:* Re: [Rtk-users] sart for planar tomosynthesis
>   Hi Robert,
> I'm not sure I understand what your tomosynthesis does. But what I would
> do is fix all geometric parameters to constant values, including the angles
> (gantry, inplane, outofplane), except SourceOffsetX, SourceOffsetY,
> ProjectionOffsetsX, ProjectionOffsetsY. In any case, I would not modify the
> object size and positions, this is a given in practice. If you want to
> cover your object, just use a larger detector.
> Good luck,
> Simon
>
> On Tue, Mar 3, 2015 at 3:56 PM, "Robert Calließ" <Robert.Calliess at gmx.de>
> wrote:
>>
>>  Hello,
>> I know that RTK's primary use is the field of a circular trajectory. But
>> if one would
>> use the the RTK for tomosynthesis the first choice's algorithm would be
>> the
>> S-ART.
>>
>> Following situation. The Focus is fixed. The object only moves in the xy
>> plane and also the
>> detectors, for instance flat panel detectors. So that a virtually line
>> would always goes from
>> the focus through the object center and ends in the detector's center.
>> I wonder how to handle the different geometry settings for this kind of
>> reconstruction.
>> What I mean is, that with a circular trajector the parameters FOD, FDD
>> and ODD always
>> remain the same because we are rotationg around an axis. But for the case
>> of planar
>> tomosynthesis the FOD, FDD will be different from the orthogonal view.
>> Usually we calculate the physical size of the object based on the FOD and
>> FDD to get sure
>> that the object fits within the cone.
>> But now for the case of a planar tomosynthesis do one need to recalculate
>> the object size
>> so that the object fits into the current cone beam or do one work with
>> the object size calculated
>> for the center view (orthogonal view, where the original FOD, FDD are
>> given) ?
>>
>> I think if I would resize the object for every viewing position this
>> would change a lot for the
>> reconstruction result as I also change the "ray paths".
>>
>> Does anyone has experiences with this kind of reconstruction ?
>>
>>
>> best regards,
>> Robert C.
>>
>>
>>
>> _______________________________________________
>> Rtk-users mailing list
>> Rtk-users at public.kitware.com
>> http://public.kitware.com/mailman/listinfo/rtk-users
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.creatis.insa-lyon.fr/pipermail/rtk-users/attachments/20150304/363580a8/attachment.htm>