search

kschan0214 / sepia

Matlab GUI pipeline application for quantitative susceptibility mapping (QSM)
MIT License
46 stars 10 forks source link

Sepia GUI malnfunction (?) #3

Closed DimitriosGk closed 4 years ago

DimitriosGk commented 4 years ago

Hello,

I have started using SEPIA GUI for susceptibility mapping. I noticed that when I choose for example Region Growing and LVB, in matlab command window I see: ---The following unwrapping method is being used: Laplacian ---The resulting field map with the following unit: Hz ---Computing weighting map...Done! ---Saving unwrapped field map...done! ---Step 2: Recovering local field... ---The following method is being used: LBV

If i runt it a couple of times without any change, I get at some point my selections printed. Is it something that occurs to other users? Maybe a re-install would solve it?

I will make sure to update, thanks in advance for any hint.

kschan0214 commented 4 years ago

Hi,

Sorry for the late reply as I was away for holidays.

The unwrapping step was designed such that when anything goes wrong with the selected algorithm, it will use the Laplacian method immediately. However, it also makes the debugging work difficult as the error message is kept.

There could be several reasons why the region growing algorithm doesn't work. For example, missing code (but I don't think is your case as you should already have the MEDI toolbox), the region growing library is not supported in the Matlab version that you use, running out of memory when running the region growing method, etc.

Could you try the following recommendations and let me know the output of them: 1) Use 'Graphcut', 'Laplacian STI suite', or 'SEGUE' (if you are using the current SEPIA master version) for the unwrapping to see if it still goes back to the 'Laplacian'? 2) What is the SEPIA version you are using? 3) What is the Matlab version you are using to run SEPIA? 4) What is the operating system you are using to run Matlab? 5) Are you using the recommended toolbox version as stated in (https://sepia-documentation.readthedocs.io/en/latest/getting_started/Installation.html)?

If you can share your one set of data (just the phase nifti file, the MAT header file and the sepia_log.m file are enough) with me (k.chan@donders.ru.nl) I can try to re-create the problem and get back to you with a more specific answer.

Hope this could help.

Kwok

DimitriosGk commented 4 years ago

Hello ,

Thank you for the thorough reply, apologies for the late answer. I was running on MATLAB 2018b, Linux Ubuntu. The SEPIA is running ok, it was my failure to read the notes and I was trying the wrong methods combination. Another problem was the fact that I used another version of the STI package, as it is not available for downloading for now and apparently this caused of course some problems as well. I have modified the code of the GUI and SEPIA accordingly in a couple of spots and now it works well.

Kind regards,

Dimitri

From: Kwok-Shing Chan notifications@github.com Sent: Monday, October 7, 2019 1:54 PM To: kschan0214/sepia Cc: Dimitrios Gkotsoulias; Author Subject: Re: [kschan0214/sepia] Sepia GUI malnfunction (?) (#3)

Hi,

Sorry for the late reply as I was away for holidays.

The unwrapping step was designed such that when anything goes wrong with the selected algorithm, it will use the Laplacian method immediately. However, it also makes the debugging work difficult as the error message is kept.

There could be several reasons why the region growing algorithm doesn't work. For example, missing code (but I don't think is your case as you should already have the MEDI toolbox), the region growing library is not supported in the Matlab version that you use, running out of memory when running the region growing method, etc.

Could you try the following recommendations and let me know the output of them:

  1. Use 'Graphcut', 'Laplacian STI suite', or 'SEGUE' (if you are using the current SEPIA master version) for the unwrapping to see if it still goes back to the 'Laplacian'?
  2. What is the SEPIA version you are using?
  3. What is the Matlab version you are using to run SEPIA?
  4. What is the operating system you are using to run Matlab?
  5. Are you using the recommended toolbox version as stated in (https://sepia-documentation.readthedocs.io/en/latest/getting_started/Installation.html)?

If you can share your one set of data (just the phase nifti file, the MAT header file and the sepia_log.m file are enough) with me (k.chan@donders.ru.nlmailto:k.chan@donders.ru.nl) I can try to re-create the problem and get back to you with a more specific answer.

Hope this could help.

Kwok

- You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/kschan0214/sepia/issues/3?email_source=notifications&email_token=AHCVNQX4XEXOXZKGZPJGC23QNMPNVA5CNFSM4IZMA4G2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEAQBHZY#issuecomment-538973159, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AHCVNQXUPIBCB4ZPTU4GWS3QNMPNVANCNFSM4IZMA4GQ.

kschan0214 commented 4 years ago

Grad to hear that!

As far as I remember, there are some discrepancies of the input variables of some functions between different STI suite versions. The STI Suite developing team seems not very responsive recently. If you need STI Suite version 3, PM me and I can send you a copy.

I'll close the issue if there is no further question. :)

Kwok

DimitriosGk commented 4 years ago

Thanks a lot for that!

I managed to get STI Suite v.3? eventually as well. Of course, you can close the issue :)

?est regards,

Dimitri?

From: Kwok-Shing Chan notifications@github.com Sent: Friday, October 11, 2019 6:44 PM To: kschan0214/sepia Cc: Dimitrios Gkotsoulias; Author Subject: Re: [kschan0214/sepia] Sepia GUI malnfunction (?) (#3)

Grad to hear that!

As far as I remember, there are some discrepancies of the input variables of some functions between different STI suite versions. The STI Suite developing team seems not very responsive recently. If you need STI Suite version 3, PM me and I can send you a copy.

I'll close the issue if there is no further question. :)

Kwok

- You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/kschan0214/sepia/issues/3?email_source=notifications&email_token=AHCVNQS5AXQ3KGUJXF4YIJTQOCUQPA5CNFSM4IZMA4G2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBASAEY#issuecomment-541138963, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AHCVNQTC2KO5LQVQ6P357LTQOCUQPANCNFSM4IZMA4GQ.