Phil Chlap

Hi @denbonte, thanks a lot for alerting us about this. I think it's really cool that your automated pipelines pick up things like this, super useful! I am very curious...

Hi @denbonte & @LennyN95, thanks so much for your support in debugging here. I have a suspicion about what could be causing this, but will have to discuss with @rnfinnegan...

Thanks for the investigation @denbonte! OK so we can rule out the multiprocessing issue... But there is something going during the atlas best registration phase... I will investigate the DIR...

I think I have found the culprit! It actually wasn't in the DIR, but in the rigid registration that is performed prior to the DIR. It turns out it was...

OK I think this is looking good, the results are now exactly the same between each run and the segmentation quality is looking good. We can run with as many...

And for those interested, this was the change that solved everything: https://github.com/pyplati/platipy/pull/198/commits/f05a5f7c73a8975af91014cdcbc421f18434cc2b

PlatiPy 0.70 is now released and available on PyPi 🚀 > That's a good point, indeed. For your interest, I found these relevant issues while surfing on GitHub and Google...

Thanks for these further insights @denbonte, always appreciated! That is disappointing to see that there are still varying results under different conditions. I agree that is most likely due to...

Hi @bwheelz36, thanks for reporting this. I imagine this is due to a new SimpleITK version. Seems like we have been discovering a few breaking changes since SimpleITK 2.3.0. I...

Hi @Ken32g, thanks for flagging this. The tool will download the models etc on first run so I this wouldn't work if there is no internet connection on the cluster....