[Rtk-users] About the Elekta Reconstruction DBF file

[Simon Rit]

The code is indeed cryptic, here is an explanation:
- if the pixel intensity is 0, then it's set to log(65536)-log(65535)
https://github.com/SimonRit/RTK/blob/master/include/rtkElektaSynergyLookupTableImageFilter.hxx#L44
This value is set when the detector saturates
- if it's different from 0 and 65535, then it's set to
log(65536)-log(65536-pixelValue)
https://github.com/SimonRit/RTK/blob/master/include/rtkElektaSynergyLookupTableImageFilter.hxx#L50
- if the pixel value is 65535, it is set to log(65536)-log(65535) (as pixel
values 0 and 1)
https://github.com/SimonRit/RTK/blob/master/include/rtkElektaSynergyLookupTableImageFilter.hxx#L56
This code is inspired by what (I think) Elekta XVI does with the raw data.
Simon

On Thu, Dec 2, 2021 at 1:02 PM 雅之 服部 <masayuki.h.0215 at gmail.com> wrote:

> Thank you for your reply.
>
> I checked RTK 3D circular projection geometry doc and understood.
>
> I have another question.
> After setting the geometry from the DBF file and reading the raw
> projection data from .his file,
> How do you convert the pixel values of the projection data to values for
> reconstruction?
> The pixel values in the .his file appear to have 16-bit values.
> Is the value of the raw data the numerator as it is, and the maximum value
> in the image as the denominator and the negative logarithm?
>
> Or is the value of raw data (probably a logarithmic system) once made into
> a linear system using the unique conversion table of the elekta system?
> I checked  rtkElektaSynergyRawLookupTableImageFilter.h  and
> ElektaSynergyLookupTableImageFilter.h  ,
>  but I couldn't understand the difference between each source code because
> I didn't study C ++ yet.
>
> Best regards.
>
> masa
>
> 2021年12月1日(水) 0:21 Simon Rit <simon.rit at creatis.insa-lyon.fr>:
>
>> Hi,
>> Yes, this mailing list is meant for this kind of questions.
>> Only this information is used, see here
>> <https://github.com/SimonRit/RTK/blob/master/src/rtkElektaSynergyGeometryReader.cxx#L72-L74>
>> .
>> To understand the meaning, you can check that it's used here
>> <https://github.com/SimonRit/RTK/blob/master/src/rtkElektaSynergyGeometryReader.cxx#L92>.
>> Yes, it corresponds to DetectorOffsetX and DetectorOffsetY in the figure of
>> this documentation <http://www.openrtk.org/Doxygen/DocGeo3D.html> for
>> each projection.
>> Simon
>>
>> On Tue, Nov 30, 2021 at 2:28 PM 雅之 服部 <masayuki.h.0215 at gmail.com> wrote:
>>
>>> Dear rtk-users,
>>>
>>> I'm masa at japan.
>>>
>>> Is it okay to email General Technical Questions about RTK here?
>>>
>>> I'm working on RTK elekta reconstruction, but I'm new to C ++, so I
>>> don't understand the specifics.
>>> I downloadedIMAGE.DBF, FRAME.DBF and looked at the contents.
>>> I understand that I am using angle information. But I didn't know what
>>> U_CENTER and V_CENTER meant.
>>>
>>> Are they the offset numbers for each angle of the detector?
>>> Also, which information is used for reconstruction from those two files?
>>>
>>> Thank you very much in advance.
>>>
>>> masa at japan
>>> _______________________________________________
>>> Rtk-users mailing list
>>> Rtk-users at public.kitware.com
>>> https://public.kitware.com/mailman/listinfo/rtk-users
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://public.kitware.com/pipermail/rtk-users/attachments/20211202/4c3a9f92/attachment.htm>