Hi,
As many other 3T Siemens sites, our center will soon transition from VE11C to XA30. We were wondering if some of you who have already transitioned could share their experience, in order to plan ahead possible roadblocks. Things like:
- Conversion to BIDS,
- Adverse effects of the enhanced DICOM format (in light of this comment on dcm2niix)
- Possible incompatibilities with popular neuroimaging tools (FSL, SPM, AFNI, fMRIprep, FreeSurfer, etc.)
- Any noticeable change on the product sequence? (eg: some features being greyed out)
- Other?
Many thanks!
Julien
@jcohenadad I am the author of the notes you refer to. The original XA10/11 images were problematic. However, Siemens engineers engaged users like me during the development of XA20. The basic XA advice still stands: users should be encouraged to export data from the console using unanonymized enhanced data - never select mosaics.
Our system is still VE11, so I have limited experience with XA30. Most of my communication with XA30 has been with sites that are unable to share data. I would suggest that when your scanner is upgraded you acquire exemplar data of all your sequences to check the conversion. I suspect that standard T1, T2, T2* (fMRI/resting state) and DWI sequences should convert fine. I have not seen ASL data from XA30, and I doubt dcm2niix will be able to extract BEP005 parameters.
If any XA30 center is interested in acquiring a validation dataset that could be publicly shared, I think that would be a really asset to the community. This would allow regression testing, similar to the other dcm_qa datasets. I would be happy to help a site set up such a repository.
1 Like
Thank you for sharing your insights @Chris_Rorden. Our upgrade is not before a few months, but I will be happy to share datasets and report any conversion/practical issues in relation to this upgrade.
We have PRISMA XA30 at our department for almost a year and we are still facing some challenges. We managed to solve some of them (the way @Chris_Rorden mentioned). However, there are a lot of bugs remaining to be solved, we are working on it and keep in touch with the Siemens service team who is trying to help us.
In our routine clinical practice, the user interface often crashes and lags, which slows down and complicates the work of our radiologists and MR technicians. Sequences cannot be uploaded from previous versions (VE11…, even from XA20) as we used to, so everything needs to be set up from the beginning, the XA30 is not compatible with the previous versions.
With FSL, SPM, etc., we only had a problem at the beginning, due to data export. However, we still have to export the data manually from the scanner because XA30 export profile still does not allow us to export the data as we would need.
We currently have the main problem with the export of MRS data and their processing, which we have not been able to solve yet. Of course, MRS analysis using syngo.MR Spectro works surprisingly great, but when we carry out more advanced analyzes, e.g. in LCModul or Tarquin, the program fails to work. If someone managed to solve the processing of MRS from XA30, we would really appreciate the information (advice) on how to do it. Our colleagues at another workplace are facing the same problems, but maybe we are missing something.
My colleague @valosekj and I are currently planning to measure some spinal cord data, we would be happy if we could share our experience and data with the community. We are open for cooperation and would be happy to share our experience and insights from our hospital.
3 Likes
Thank you so much for sharing your valuable experience @Dominik_Vilimek, and sorry to hear you went through this trouble 