SimpleITK==2.0.2

[patricio-astudillo]

Dear @FabianIsensee,

if I use versions higher than 2.0.2, I receive the following error: ITK only supports orthonormal direction cosines. No orthonormal definition found. This is also mentioned in (MIC-DKFZ/nnDetection#24).

Could please undo your change in the requirement.txt file please? I have corrected this in the past and now I receive the same errors as before.

If you want I can push this change also, but this seems a bit back and forth to me.

Best regards,
Patricio

[FabianIsensee]

Hi, can you please specify what dataset you are having problems with? I just upgraded my SimpleITK to SimpleITK==2.1.1 and have not had any problems with the commands I tested.
Best,
Fabian

[patricio-astudillo]

Hey Fabian, It is a private dataset from a hospital. However, I don't think it is dataset related. However, I don't want to block updating any versions of SimpleITK but I also want to make sure that others do not run into the same troubles that I have had. How shall I proceed to investigate this? I can do a pip show to list all my versions and provide an overview of my system (os, etc...)? Thanks in advance for your help, Patricio Op do 20 jan. 2022 om 08:33 schreef Fabian Isensee ***@***.***

…

: Hi, can you please specify what dataset you are having problems with? I just upgraded my SimpleITK to SimpleITK==2.1.1 and have not had any problems with the commands I tested. Best, Fabian — Reply to this email directly, view it on GitHub <#25 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB4QPT6FP7FAX62Q73UXMMLUW63C7ANCNFSM5MEX3L5A> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[Xtaltec]

Had the same issue with a private dataset and it was solved by downgrading to SimpleITK 2.0.2.

[machur]

We've encountered the same problem on a private dataset. Only couple of MR studies were affected for hundreds that were processing. It doesn't seem like an easy case to reproduce, especially considering the fact that all failed cases looked quite "orthogonal" by visual inspection. It's probably some corner case of affine matrix processing by SimpleITK.

[machur]

@patricio-astudillo I upgraded SimpleITK from 2.1.1.2 to 2.2.1 and it has worked with the new version.

[FabianIsensee]

SimpleITK has changes the way it works with geometries. It's unfortunate and causes quite a few issues. Changing the SimpleITK version will resolve this, thanks @machur