Issue opening .DCM files

[Alireza Meghdadi]

I have an issue opening a DICOM image with the .DCM extension. The following window shows up:

Capture_3.PNG (33.57 KiB) Viewed 524 times

UPDATE:
I tried creating a new project. Upon adding the DICOM file, the terminal showed a series of similar outputs corresponding to "Reader 0" to "Reader 5", with the latter giving an error:

Capture_4.PNG (44.91 KiB) Viewed 516 times

[David Parker]

Hi Alireza,

There is a known problem reading in some DICOM image data (see https://github.com/SimVascular/SimVascular/issues/510). The fix for this will be in the next SV release scheduled for the end of this week. As a work around you can convert the DICOM files to a VTK VTI file using vmtk.

Cheers,
Dave

[Alireza Meghdadi]

Thanks a lot! I will give that a go.

[Alireza Meghdadi]

Do you happen to know of good vmtk tutorials? their instructions are very vague and I am not a professional programmer; I have limited knowledge enough to get me started.

[David Parker]

Hi Alireza,

Have a look here http://www.vmtk.org/documentation/getting-started.html. It describes the vmtk command to convert a DICOM image volume into a VTK VTI format file. SV can read in the VTI file. I've tried this and it works no problem.

You can download vmtk from here http://www.vmtk.org/download. Use the executable installers, not conda (it will drive you insane).

Cheers,
Dave

[Alireza Meghdadi]

Hi Dave

I tried VMTK, but it fails to convert my DICOM image. I get errors and I cannot find the root of the problem. My DICOM images are totally fine, I can view them in MicroDicom. I guess I will be looking out for the next SimVascular update hoping it can import my images.

FYI, the error I get in vmtk (PypePad):
{
ERROR: In ..\Common\DataModel\vtkDataSet.cxx, line 435
vtkImageData (000001EDEE05D740): Point array ImageScalars with 3 components, only has 1 tuples but there are 67502080 points

ERROR: In ..\IO\XML\vtkXMLStructuredDataWriter.cxx, line 388
vtkXMLStructuredDataWriter (000001EDEDADE630): Input is invalid for piece 0. Aborting.

ERRORL In ..\Common\ExecutionModel\vtkExecutive.cxx, line 784
vtkCompositeDataPipeline( 000001EDEC257440): Algorithm vtkXMLImageDataWriter(000001EDEDADE630) returned failure for request: vtkInformation (000001EDEE05CA00)
}

[David Parker]

Hi Alireza,

Sorry vmtk didn't work. It uses GDCM to read DICOM images, guess it doesn't handle your image format. Can you use MicroDicom to export to VTI format?

I looked at the VTK source to see what the error might mean, seems that you have image data with three components but only a single value, I can only wonder what that might mean! This is not PC MRI data is it?

There is a new SV Windows installer on SimTK. Please install and see if that works.

Cheers,
Dave

[Alireza Meghdadi]

Hi Dave,

Thanks a lot for taking the time to attend this issue.

I am not aware of the type of DICOM I have acquired. I had a large array of files (I assumed they were slices) that I imported into MicroDicom and exported a .DCM image which is the file I have been trying to import into SimVascular. Strangely, the very first time I tried importing it right after I installed SimVascular for the first time, the import worked successfully. However, when I tried to save the project, the software was not responding so I closed it. After that, all imports have been failing. I have also tried the July 10th release but no luck.

Regarding your interpretation of the error message, only one thing comes to mind. When I imported the directory of slices into MicroDicom, there appeared to be three different sets of images. One raw data, One with arterial contrast and one with venous contrast. I am interested in the venous structures so I only exported that set of slices. Maybe this can help you make some sort of sense from the error message!

Thanks a lot,
Ali

[David Parker]

Hi Ali,

So you are creating a new project and trying to import the data and that is failing? Make sure that you remove all remnants of the previous SV release.

I am guessing that the MITK framework that SV uses to read DICOM files does not understand the single multi-frame DICOM file, or maybe it is the raw data that is confusing it. We typically read in DICOM images as a series of files. Could you convert just the venous contrast image data to DICOM and see if that works?

Cheers,
Dave

[Alireza Meghdadi]

Hi David,

That is what I have done so far. So after importing the series in MicroDicom, I export the Venous series as a separate DICOM, then I try to import this into SV and I get the error as seen in the image I uploaded at the top of this discussion.

Thanks,
Ali