This is an issue for scanners with an even number of tangential positions because the fan data does not account for the min_tangential_poss() element when applying norm factors. The resulting efficiency/ normalization values in min_tangential_poss() positions will be 0 or +inf, which will lead to numerous problems.
This is an summary of the issue discussed in #1499 , specifically from https://github.com/UCL/STIR/pull/1499#issuecomment-2330236446 onwards.
This is an issue for scanners with an even number of tangential positions because the fan data does not account for the
min_tangential_poss()
element when applying norm factors. The resulting efficiency/ normalization values inmin_tangential_poss()
positions will be0
or+inf
, which will lead to numerous problems.The fix would be to extend the fan data to allow for an even number of entries, but this code issue is spread throughout ML Norm (https://github.com/UCL/STIR/pull/1499#issuecomment-2332808262)