search

jfcabana / omg_dosimetry

Optimized Multichannel Gafchromic Dosimetry
https://omg-dosimetry.readthedocs.io/en/latest/
MIT License
8 stars 2 forks source link

Normalization Order/Timing (during tiff2dose as opposed to analysis stage) #56

Open petruong opened 2 weeks ago

petruong commented 2 weeks ago

Hello there,

tiff2dose module is an interesting module in itself as it contains plenty of relevant detail of the scan converted into dose (by channel, mean, optimized, other noise evaluation formats, etc.). However, it can also be used as a simple tool to output the dose map for scanned film.

At this stage, it's possible to run tiff2dose module with scanned film and the corresponding LUT path, which generates the resultant dose map. Typically, this quickly gets utilized in the analysis stage in order to compare with a reference dose plane. Well, what happens if I simply wanted a dose map just to view the dose gradients/field shape (and not have a comparison with a reference dose plane)? The current optimized dose map (tiff2dose stage) is not yet normalized (or taking into account future implementation of the two-step zero reference film normalization). Would it be relevant to have the normalization steps (found in the analysis module) be transferred to the tiff2dose module and have it save the optimized dose map file and include the option to save as a dose map with normalization factor applied (i.e., 10x10 film_opt_normfacteur0.99). This way, if you wanted to view the dose map, at least it'll have its corrections from the beam output of the day, aging of the film (both to account time between irradiation and box age). The normalization/zero dose film impacts the scanned dose map anyways and it's not necessarily needed at the analysis module stage (i.e., doesn't apply to the reference dose plane nor gamma analysis step).

The major issue in all of this will be the execution/philosophy of the QApatients file/script as the timing to generate and then use the created dose map will change since it'll now include the normalization/zero dose film during tiff2dose module.

Anyways... Is it worth it? If so, how do we want to go about it?

jfcabana commented 2 weeks ago

Sounds quite logical I think. Changing the script files won't be such an issue. You could create a new branch implementing these changes. Maybe it could be made at the same time as the two-point correction method?