long object artifact

[Zhentao-Liu]

I noticed an interesting phenomenon.
When I perform FBP reconstruction with a full 360° (1200 frames) scan, both the left (L) and right (R) sides exhibit long-object cone-beam artifacts.
However, when I reconstruct using a short 240° scan (>180° + fan angle (49.5°)), the artifacts on the left (L) side are reduced, while those on the right (R) side become more severe.
Why does this happen?

[kylechampley]

LEAP uses zero-th order extrapolation when performing the backprojection step as a part of FBP reconstruction. The amount of extrapolation used will depend on the angular coverage. Another way to perform FBP reconstruction is the following:

leapct.filterProjections(g)
leapct.weightedBackproject(g,f)

if you don't want LEAP to do this extrapolation you can just tell it to use the standard backprojeciton, which for cone-beam, the only difference is the extrapolation.

leapct.filterProjections(g)
leapct.backproject(g,f)