Unable to execute part 0 of module dvm1 (DICOMReader) ERROR

[fxmech]

Hello Devide group! I'm a new member and this is my first post.

 

 

Does anyone know if there is a way to correct a Dicom dataset that is yealding this error.

I have a collection of MRI studies...  some are read by Devide just fine and others give me this error log:

 

 

21:33:09: Traceback (most recent call last):
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\module_manager.py", line 879, in execute_network     self._module_dict.values())
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\network_manager.py", line 46, in execute_network     self._devide_app.scheduler.execute_modules(sms)
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\scheduler.py", line 698, in execute_modules     self.get_scheduler().execute_modules(scheduler_modules)
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\scheduler.py", line 569, in execute_modules     mm.execute_module(sm.meta_module, sm.part)
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\module_manager.py", line 848, in execute_module     meta_module.execute_module(part, streaming)
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\meta_module.py", line 358, in execute_module     self.instance.execute_module()
21:33:09:   File "C:\Program Files\DeVIDE-RE\devide\modules\readers\DICOMReader.py", line 167, in execute_module     raise RuntimeError(msg)
21:33:09: ModuleManagerException: Unable to execute part 0 of module dvm1 (DICOMReader): DICOM IPP sorting yielded incorrect results.
21:33:09: Unable to execute part 0 of module dvm1 (DICOMReader): DICOM IPP sorting yielded incorrect results.

 

I tried reviewing the images in the DICOMBrowser but did not notice anything unusual in the series of slices....  maybe I just don't know what to look for.

Is there a way to correct the set?  I tried loading the first 3 slices and it worked fine but with slice 4 the error reoccurs on this particular set... tried deleting slices in DICOMReader but that did not help.  Is this set unusable or is it possible to correct it somehow???

 

 

 

 

[Martijn Steenwijk]

Reading DICOM can be tricky due to different vendors, etc. An aproach to circumvent the problem is convert the dicom to nifti (with eg dcm2nii), then load the data using ITKReader and subsequently use the ITKtoVTK-module to convert to VTK.

Best,

Martijn

 

 

 

 

--
You received this message because you are subscribed to the Google Groups "devide-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/devide-users/-/S_xDa7UCtXUJ.
To post to this group, send email to devide...@googlegroups.com.
To unsubscribe from this group, send email to devide-users...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/devide-users?hl=en.

[Charl Botha]

Dear fxmech,

What Martijn suggests is certainly valid. 

The error message you're seeing means that the spatial positioning information included with the MRI slices does not line up, i.e. they do not stack neatly. This could be because the MRI scanned using a non-rectangular geometry, or that there's a single slice in there (usually at the start or the end) breaking the sequence.

You could also try (as an experiment) to load just the middle 5 slices, or the 2nd to the 6th.

In any case, for extracting skull data, CT is a much better bet.

Good luck,

Charl

--
dr. charl p. botha :: http://graphics.tudelft.nl/CharlBotha
work: c.p....@tudelft.nl personal: cpb...@cpbotha.net

[fxmech]

Thanks, 

I was able to get the same files to load using the older dicomRDR but with no luck isolating the skull info.  Note that the reason is that it was a brain MRI...  I am planning on doing a CT to make it easier to extract the skull info.  A CT tech I was talking to mentioned that I should request a HELICAL CT with no contrast to get the best results.  Does this sound right??? 

My goal is to build a 3D mesh of my skull and place it on a 3d canvas along with a 3D mesh of my face.  This should allow me to work within the negative space in between the two layers.  The goal is to use this data to build an animatronic head truer to actual human anatomy and actuated in a way resembling the motion of facial muscles.... a lot more servos but hopefully a lot more realism.

I am using myself as the subject because I am always available for reference ;)

I just need to get the skull 3d model and I'll be in good shape....

[Charl Botha]

On Thu, Aug 23, 2012 at 7:06 AM, fxmech <fxm...@cammotion.net> wrote:
> I was able to get the same files to load using the older dicomRDR but with
> no luck isolating the skull info. Note that the reason is that it was a
> brain MRI... I am planning on doing a CT to make it easier to extract the
> skull info. A CT tech I was talking to mentioned that I should request a
> HELICAL CT with no contrast to get the best results. Does this sound
> right???

It does, but it's not wise to request a CT purely for this purpose.
The radiation dose is significant, so a CT should only be made if
clinically useful.

> I am using myself as the subject because I am always available for reference
> ;)
>
> I just need to get the skull 3d model and I'll be in good shape....

See my tip above. I would definitely start with one of the datasets
that have been made available by osirix instead.

[fxmech]

I have heard about the risks from most of the techs I've talked to And honestly I still dont know that I can get a dr to ok the scAn for research purposes. I do believe that there is huge potential for this technogy outside of the movie industry infact the. Chances of applying this to film is small because of the complexity and cost/time factors involved. I just feel that the health risks of working with the chemicals involved is probably much higher than the ct scan itself....
Having said that I am open to other options... I thought that an alternate option could be maybe to hire someone with more devide expertise to extract the data from the brain MRI I currently have... Another option is to get a 3ds max artist to modify a heretic skull mesh and reform it to closely match the shape of my skull using the MRI data as a template....
If any group member is interredted in helping me jump this hurdle please let me know cost. Time is not a factor as this is a long term project.