radionics error about raidionics

[dbouget]

Hi @deem32,

It appears that no contrast-enhanced T1-weighted MR scan has been loaded for the current patient, and the Brain segmentation model expects to find one. I assume you tried running a custom action for segmenting the brain on MR scans available at timestamp T0?

You have to make sure that the correct Sequence type is selected for each MR scan you have loaded (in the right-hand side panel showing the Data). If you have multiple MR scans (with different sequence types) loaded for one patient, you might want to try with simply loading a single MR scan (i.e., the one you want to segment) and see if it works there.

If you haven't already, you could have a look and run the simple example to make sure everything is set up properly on your machine. Simply go through the following tutorial.

Let me know if you succeeded or if you have further info to provide in order to narrow down the issue!

[deem32]

[image: image.png] David Bouget ***@***.***>, 12 Şub 2024 Pzt, 17:26 tarihinde şunu yazdı:

…

Hi @deem32 <https://github.com/deem32>, It appears that no contrast-enhanced T1-weighted MR scan has been loaded for the current patient, and the Brain segmentation model expects to find one. I assume you tried running a custom action for segmenting the brain on MR scans available at timestamp T0? You have to make sure that the correct Sequence type is selected for each MR scan you have loaded (in the right-hand side panel showing the Data). If you have multiple MR scans (with different sequence types) loaded for one patient, you might want to try with simply loading a single MR scan (i.e., the one you want to segment) and see if it works there. If you haven't already, you could have a look and run the simple example to make sure everything is set up properly on your machine. Simply go through the following tutorial <https://github.com/raidionics/Raidionics/wiki/Frequently-Asked-Questions-(FAQ)#0-verification-test> . Let me know if you succeeded or if you have further info to provide in order to narrow down the issue! — Reply to this email directly, view it on GitHub <#68 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BGC2BKH7A6M4CBRIB75TL73YTIRCZAVCNFSM6AAAAABDEFZI5GVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZYG44DAMZWGQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[deem32]

Following your suggestion, I applied the test data and the discount steps exactly. Unfortunately, I was not successful again. I added the error code, I would be happy if you help. dr deem ***@***.***>, 12 Şub 2024 Pzt, 16:43 tarihinde şunu yazdı:

…

[image: image.png] David Bouget ***@***.***>, 12 Şub 2024 Pzt, 17:26 tarihinde şunu yazdı: > Hi @deem32 <https://github.com/deem32>, > > It appears that no contrast-enhanced T1-weighted MR scan has been loaded > for the current patient, and the Brain segmentation model expects to find > one. I assume you tried running a custom action for segmenting the brain on > MR scans available at timestamp T0? > > You have to make sure that the correct Sequence type is selected for > each MR scan you have loaded (in the right-hand side panel showing the > Data). If you have multiple MR scans (with different sequence types) loaded > for one patient, you might want to try with simply loading a single MR scan > (i.e., the one you want to segment) and see if it works there. > > If you haven't already, you could have a look and run the simple example > to make sure everything is set up properly on your machine. Simply go > through the following tutorial > <https://github.com/raidionics/Raidionics/wiki/Frequently-Asked-Questions-(FAQ)#0-verification-test> > . > > Let me know if you succeeded or if you have further info to provide in > order to narrow down the issue! > > — > Reply to this email directly, view it on GitHub > <#68 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BGC2BKH7A6M4CBRIB75TL73YTIRCZAVCNFSM6AAAAABDEFZI5GVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZYG44DAMZWGQ> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> >

[dbouget]

Hi @deem32,

It seems the last upload of your error code did not succeed, the image itself does not show.

The test data provided inside Raidionics should enable running the preoperative segmentation process without any error.
At the moment, I am not able to reproduce your error. If you have the possibility, you can also send me the MRI scan you tried using initially, maybe it will help me figuring out the potential problem.

What you can also verify inside the settings (Settings > Preferences > Inputs/Outputs), is that the Use manual MRI sequences box is ticked. The sequence type that you manually associated to each MRI scan that you loaded will then be used. Otherwise, a classification model will be run automatically, but the model hasn't been thoroughly assessed and cannot be highly trusted.