[Rtk-users] Forward projection to generate realistic fluoroscopic images

[Simon Rit]

Hi,
1/ No, it's probably not the best input for generating DRRs. I would
suggest to put air at ~0 by adding -1024 to all pixel values.
2/ The output are line integrals. Fluoroscopic images should follow the
Beer Lambert law so the best would be to chose take the exponential of the
opposite. However, you should then be careful in your conversion of the HU,
having water at 1024 would not be numerically complicated. So multiply the
HU+1024 values by a scalar which would put water at a reasonable value,
e.g. 0.01879/1024, where 0.01879 / mm is the linear attenuation coefficient
of water at 75 keV.
I hope this helps,
Simon

On Mon, May 31, 2021 at 10:53 AM Kobe Bamps <kobe_bamps at hotmail.com> wrote:

> Hi all,
>
>
>
> I am using the JosephForwardProjection of RTK through python wrapper. I am
> trying to generate realistic fluoroscopic images from a chest CT-volume.
> I’ve setup the image spacing and pixels based on the dimensions of the
> detector of a c-arm. I can generate projections over different angles.
>
>
>
> Question 1: Should the intensity values of the input volume be expressed
> in Houndsfield Units? In other words, should the intensity values range
> between -1024 HU and 3071 HU?
>
>
>
> Question 2: What is the best practice to generate positive projection of
> the CT-volume instead of a negative projection (like on x-ray display
> devices or fluoroscopes. Bones are black, air is white)?
>
>
>
> It would be greatly appreciated if anyone could answer on my questions.
>
>
>
> I apologize if I am vague on any points or if anything is unclear, please
> let me know.
>
>
>
> Thank you,
>
> Kobe
>
>
>
>
> _______________________________________________
> 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/20210531/ade87d39/attachment.htm>