search

UCL / STIR

Software for Tomographic Image Reconstruction
http://stir.sourceforge.net/
Other
111 stars 93 forks source link

BinNormalisation::apply should use timing info from ExamInfo #340

Closed KrisThielemans closed 3 years ago

KrisThielemans commented 5 years ago

BinNormalisation::apply uses explicit timing arguments (only used for dead-time). we should get these from the ExamInfo instead.

KrisThielemans commented 4 years ago

postpone till #44 done, as there are changes there to the norm classes

KrisThielemans commented 4 years ago

seems best to pass exam_info to set_up

danieldeidda commented 4 years ago

seems best to pass exam_info to set_up

Shall we have a exam_info member in BinNormalisation?

KrisThielemans commented 4 years ago

would then have to be, yes. The point is that we shouldn't have to pass timing explicitly in all the apply/undo calls. (We'd have to allow it in any initial step, as otherwise we get into backwards compatibility issues and trouble with SIRF and STIR versions).

However, note that this is the longer term solution as we discussed. I think we should only tackle this after merging #564 for instance.

danieldeidda commented 4 years ago

with "any initial step" you mean something like set_up(). right?

KrisThielemans commented 4 years ago

Depends where we are on the path. I think we can resolve #446 (with decay) first without working on this one.

danieldeidda commented 4 years ago

BinNormalisationWithCalibration is ready but it only includes calibration factor and branching ratio. Didn't we say we need to separate decay correction for SPECT and PET?

KrisThielemans commented 4 years ago

yes. sure.

danieldeidda commented 4 years ago

The part for SPECT is already done shallI leave it there or shall I create something like BinNormalisationDecayRotatingScanners and BinNormalisationDecayStaticScanner (maybe static is not the best world?) This can directly be derived from BinNormalisation I would think.