Dear developer
Recently, some strange things have happened during my reconstruction attempts.
In my understanding, With ac factor and norm factor in both CDF and Normalization file is the right way to reconstruct pet image. but i’ve got confused in recent results.
I was tried to reconstruction a GATE-simulation cylinder (included only true and scatter events) with scatter rate from opensss. here is the comparison of reconstruction image.
there are 4 setting in my attempts:
- CDF_wtof + Mumap-AC = CDF([time, tof, id1, id2]) + mumap.hdr
- CDF_wtof_wscf + Mumap-AC = CDF([time, scatter_correction_factor, tof, id1, id2]) + mumap.hdr
- CDF_wtof_wscf + Norm_wacf = CDF([time, scatter_correction_factor, tof, id1, id2]) + normalization file with attenuation correction factor and all 1 norm factor in it.
- CDF_wtof_wscf_wacf + Norm_wacf = CDF([time, attenuation correction factor, scatter_correction_factor, tof, id1, id2]) + normalization file with attenuation correction factor and all 1 norm factor in it.
The events num and colorbar of image is totally the same, but reconstruction with setting 4 show that the central region of the cylinder is lower than expected. I suspect it might be because my scatter rate is being additionally attenuation-corrected once more inside CASToR, so I tried dividing the scatter correction factor by the attenuation correction factor before reconstruction (CDF_wtof_winvacscf_wacf + Norm_wacf) and obtained results similar to those from other settings.
Similarly, I also conducted a test on the random rate and obtained results similar to those of the scatter rate.
I would like to know whether this result is caused by CASToR applying an additional attenuation correction to the random and scatter correction factors, or is this actually the correct result, and my understanding is incorrect?
I would greatly appreciate your reply.
Best regards,
Sean