We are trying to use a tractography that we have generated using MRtrix for a surgical case. We would like to load the tractography into the Medtronic neuronavigation suite, but we are having some issues while converting the .tck to a .dcm that could be read by the neuronavigation.
We know that Karawun could do this conversion for a Brainlab suite, but we are not sure that this would work on Medtronic. Is there anyone who has had the same problem? Otherwise, how could you use a .tck during Medtronic neuronavigation?
We’ve done some limited checks of this. The short answer is that at the at moment we don’t think the dcm format produced by karawun will allow any version of medtronic stealth to display streamlines.
The longer version - StealthStation V7 has no capability to display streamlines in the way mrview/brainlab to (3d ball of yarn view). StealthStation V8 apparently does have this capability, but we haven’t had a chance to test it. We strongly suspect that the dicom format it will support, if any, will be the newer tractography result format. Karawun uses a predecessor format ( based on surface segmentation), but we hope to update in our next release.
Hello, this is the workaround I am using which allows to use Mrtrix tractograms in the OR using the S7 (and the S8 of course):
output and convert Mrtrix’s CSD peaks-derived whole brain tractogram (in tck format) into a trk file using nibabel’s nib-tck2trk module, registering with your anatomical reference image;
do the same conversion with your tract of interest;
load the tractogram trk file in DSI Studio, then add the tract;
load your anatomical image of choice and leverage the excellent registration function of DSI Studio (it’s automatic but I usually do a couples of refines playing around with the registration options until I’m satisfied and I see an optimal registration between the whole brain peaks file and the anatomical image);
use the Print tract on image function;
export it as nifti;
load the export in Slicer and convert it to a DICOM series.
You then load the series to S7 as you would normally do, merge it with your reference and build a 3D model to highlight the tract with your color of choice. It is far better if you use a T1 without contrast as the tract is printed in white. If you happen to have a T1 with contrast you can lower the threshold in DSI Studio before printing the tract so the contrast medium appears grayish compared to the actual tract and you can then adjust the model threshold in the S7 to isolate the tract without contrast artifacts.
Personally, the first two steps are embedded in my workflow script which is actually fully automated, as I’m processing DICOMs to BIDS-ready nifti files, pipe these to QSIPrep, and finally process its output with TractSeg (at least when I want a CSD) to obtain tracts.
I feel like DSI Studio is superior to other visualization softwares (including Mrview, Mi-Brain and TrackVis to mention some I’ve been testing) as it allows for easy tract handling and editing in the 3D view.
The stealth reads nifti. If not, load the nifti in 3D slicer and go to dicom conversion and create a dicom sequence from your nifti volume (easy and intuitive). Then load it into the stealth
I’ve noticed DSI Studio started saving nifti with the wrong level/contrast, hence I’ve dug deep into finding a solution which outputs tracts in DICOM format properly coregistered to the anatomical reference:
instead of using the Mark tract on slices function, go to Tracts > Save tracts > Save current tract in slice space and save your tract. Ensure the .nii extension is selected
open the anatomical reference using DSI Studio’s Tools > O1: View images
Header > Set transformation, copy the matrix info
open the tract of interest in NIFTI using the same tool as before
Header > Set transformation, paste the matrix info and hit Ctrl + S to save.
load the tract NIFTI in Slicer and convert it into a DICOM series.
Voilà ! You will also need to convert the anatomical reference to DICOM as well using Slicer’s converter. This will serve as your mask, and you will then need to use the “Pre-merged” function on the stealthstation when registering tracts.