search

gschramm / pyapetnet

a CNN for anatomy-guided deconvolution and denoising of PET images
https://gschramm.github.io/pyapetnet/
MIT License
13 stars 4 forks source link

Generating target data used in training #6

Open HyunKyungHan opened 1 year ago

HyunKyungHan commented 1 year ago

Hello, I want to train pyapetnet with my own data, but am having trouble generating the ground truth data. I know you used in-house software to reconstruct it. Is there any way that I can access this software or get ideas about how to generate good quality ground-truth data?

Thanks.

gschramm commented 1 year ago

Unfortunately, using our in-house software is possible, but non-trivial.

I guess you want to run Bowsher recons on data from your PET/MR system. Which system do you have?

HyunKyungHan commented 1 year ago

Thank you for replying! I have data acquired from PET/CT and MR system. By the way, is the software incorporated in the PET/MR system? I thought I could use reconstructed PET/CT images as the input for the software.

gschramm commented 1 year ago

Which PET/CT model + vendor are you using?

gschramm commented 1 year ago

There is a WIP package for the Siemens Biograph mMR (but not for Siemens PET/CTs or other PET/MRs)