Exporting ROIs into NIFTI format, PLUG-IN for ROI Export Automation?

[bioni...@gmail.com]

Hello Horos Team,

     Our team has a bevy of ROIs that we would like to export to a NIFTI (or any other ROI format if NIFTI is not possible) for use in algorithmic model development. Our ideal way of going about this task would be the utilization of a Horos Plug-In to convert our ROIs into the NIFTI format. We also have linear measurements along with polygon ROIs that we would like to filter out during the NIFTI conversion process. Can you help us or direct us in the proper direction for this task? This plug-in would be of high impact to our neuroimaging workflow and ability to reach vital research deadlines. You guys were so helpful in resolving our previous barriers to further Horos implementation so I thought I'd bring another challenge to your team. Thanks!

-Hasan@BIOS

[H Air]

What is NIFTI?

2018年9月15日土曜日 3時44分19秒 UTC+9 bioni...@gmail.com:

[Soren Christensen]

Hi,

 This plugin will binarize your ROI and store as DICOM (1/0 voxels). From there you could just convert to nifti in the same way you convert your images to nifti.

https://github.com/sorenchr2011/voxvol

I don't know if linear measurements would also be rasterized by the plugin. You'd need to change the code to ignore the linear measurements if so.

Soren

--
You received this message because you are subscribed to the Google Groups "Horos Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to horos-projec...@googlegroups.com.
To post to this group, send email to horos-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/horos-project/8dab554c-a83e-4a64-b726-78bfb09ba3cf%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[bioni...@gmail.com]

Hey Soren,

    This plug-in works really well for us :). We are experiencing one minor inconvenience in that the plugin-generated binary mask series does not align with the original CT within a program such as Mango or MRIcronG when attempting to use the mask series as an overlay. We have gotten around this issue by re-exporting that plugin-generated mask series, using steps similar to reconstructing a thin slice CT into a thicker slice CT, within Horos , which we then use due to its' ability to align with the original CT. Have you encountered this issue and have you ever attempted to fix this bug within the plugin's code? I understand the plugin's main purpose to be ROI volume generation, which it does well, but the added bonus of ROI binary mask generation with series export is an immensely helpful tool that also comes with the package. Any suggestions for us?

[Soren Christensen]

Hi Hasan,

 Great to hear that it is useful for you!

In Osirix the produced DICOM syncs well spatially and all the positional DICOM elements are identical - so I don't believe there's mistakes in the written orientation.

My best guess is that the behaviour you described is due to the fact that the binarized mask is exported as Secondary Capture (SC). It is possible that your software won't sync to SC modality. It should be a real simple change in the plugin to change the modality to CT, but I have not tried it. Ideally I think it should simply copy the modality of the input, but for some reason the output DICOM becomes SC modality without explicitly setting this.

Perhaps someone here knows if  [dicomExport setSourceFile: [curImg srcFile]]; implies SC modality?

Until then, I think you can manually set modality to CT and see if this solves your issue.

If you find a solution please share and we can modify accordingly.

Thanks!

Soren

To view this discussion on the web visit https://groups.google.com/d/msgid/horos-project/1182a671-9efd-4b09-b081-2f73de7a4521%40googlegroups.com.