search

jacg / petalo

PET scanner image reconstruction
1 stars 4 forks source link

Fix speed of light #24

Closed jacg closed 2 years ago

jacg commented 2 years ago

The speed of light was out by a factor of 10^9. I'm still puzzled by why this does not seem to have had grave consequences for reconstruction.

jacg commented 2 years ago

Fixing the speed of light (factor of 10^9) seems to give a slight improvement in FOMs:

tail -1 deleteme-before-fixed-c/foms
04-20   53.7   60.3   66.9   73.1   70.0   75.5            26.1   24.7   24.4   24.4   24.0   23.4           304.0  325.2  226.3  331.0  249.8  270.2 

tail -1 deleteme-fixed-c/foms
04-20   56.1   61.5   68.0   74.5   71.4   77.4            21.7   20.9   19.9   19.9   19.8   19.6           378.5  358.4  256.7  365.5  272.6  298.5

(The groups are CRCs, background variability and SNR.)

This used an attenuation image which was made before the fix, but off the top of my head I can't think of a reason why that would depend on c.

jacg commented 2 years ago

This fixes a problem in vislor where the TOF option appeared not to work, because whatever TOF you specified was interpreted as 10^9 times that value, leading to a gaussian which was essentially constant across the whole FOV.