MITK Diffusion 2012 does NOT crash..

Reproduced on:

I put importance on "high" and included the 2012-12 release as target milestone, since

This is probably an IVIM-related bug

A further test with MITK-2012.06.99_r5d2306-Linux:

asdf

I already took a quick look, deletion of AppData/DKFZ does not help ... I am not sure, what we are facing here.

the changes seem to have broken the input data selection process. once a mask was selection, it is hardly possible to deselect it and continue voxel-based analysis.

Hi Peter, the PV-analysis works either with DTI or scalar images. DWI images are treated as scalar so far as I know.

Peter, Iggy, is this on our radar for the release?

This seems to be an encoding issue: "none (2,5,0,0) (0,2,5,0) (0,0,2,5)"
should be "none (2.5,0,0) (0,2.5,0) (0,0,2.5)"

I just reproduced the bug with the current installer, fiber extraction crashes on both, circular and polygonal extraction.

Any new insights here?

the button does not produce any output. I tried it with a DTI image. Nothing happens when pressing the export button.

I don't know about special characters, but it was very long ;)

cool ;)

Hi Stefan, could you comment on the status of this bug? This could be usefull for our release, or at least bugfix release.

Hi Martin,
thank you for reporting this.
Tom, have you encountered that bug already?

Yes, this would be nice. I wonder if it is stable enough to just include it in a "bugfix" release... or if we have to wait for a real release.

Außerdem: Im info-screen zur TBSS FSL anstatt FLS

Weiß nicht, sollten wir bei den Views ein Textfeld mit Kommentar einfügen, welche Art von Daten benötigt wird?

Außerdem: stellen wir Testdaten online zur Verfügung?
"IVIM Datensatz liegt nicht vor, Rohdaten zum Dicom import testen liegen nicht vor"

this does not seem to work, loading adriane.

Einige Dateien sind sehr groß:

Caspar, I think you have to fix this (is unique to release branch), just change the index in the qt-designer ui-file on the bottom of the GUI at advanced settings.

thomas tbss images I think

I think we should deactivate the option in the app. This is too close to release now. lets open a new bug.

organize buttons, get rid of numbered buttons.

the option "regularized IVIM" should be deactivated in the GUI. its work in progress.

bug-status = fixed?

we should be fine I think..

Maybe MBI and MINT should be acknowledged as a group, the diffusion team and Marco/Sascha should be listed.

OK, thanks, the cmake tips solved this.

well, at least under MY linux ;) I am not sure, if I am missing anything here with my configuration..

maybe this is relevant information for this problem?:

Still does not work under Linux.. even though MITK was build before :(

(In reply to comment #12)

If you do a search of your binary tree, do you have .qch files generated for
all those diffusion imaging .dox files? Do you have a DiffusionImaging .qhc
file? Or are they just missing?

I am experiencing the same as Thomas.

Hi Iggy, this could be a different problem with the DWI datatype, lets file a new bug for that one.

You are right, together with the image the visualization works.

this is a geometry issue, zooming out a lot lets the dataset appear.

what are we going to do here?

I think, the code-review should not cover too much detail, mainly the basics:

Strange indeed.. raw and header files in the same folder? Does the link to the raw file in the header contain any slashes? Or is it just the pure file name?

Anybody has a clou, what this is about?

ups, I did not see your comments.. forget mine.

Yep, this is fixed.

I think this is a feature of general interest. But this is a long discussion;)

crossing my fingers;)

Iggy, could you coordinate?

Major issues remaining in the bug-list:

? Why not?

I will fix it, looks like a nasty stack corruption in ODF mapper.

Hi Caspar, we should now get rid of master view, and show only relevant diffusion related perspective icons in the toolbar.

Some discussions later, here goes a new plan:

Peter reported the same behaviour when testing the diffusion windows installer. Might this be a diffusion related bug?

The app could contain several perspectives for the different tasks:

New plan for the master view to be realized in T9156

let's talk about this, I implemented something similar for the atlas app.

cubic interpolation (faulty or at least not intuitive)

nearest interpolation (correct)