Tobias Block
@tblock79
I am currently using Qt 5.6.3 to build the releases
BlakeGFitch
@BlakeGFitch
So I'll investigate assembling a build environment tomorrow
Do you know if the siemens reconstruction can be done outside of the scanner system? I'm asking this thinking about reproducible science. I'd like to capture the ability to reproduce the .dat -> dicom independent of the siemens system.
The idea is to be able to reproduce everything from the raw data forward and be able to archive that capability.
Tobias Block
@tblock79
Unfortunately not. You would have to develop your own reconstruction methods for that, e.g. using Gadgetron or BART. Unfortunately, there are also certain system-specific things like the Distortion Correction (which requires knowledge of the exact gradient fields) that Siemens is keeping a secret
BlakeGFitch
@BlakeGFitch
Interesting
Not so supportive of reproducible science tho.
How do you deal with that in your work?
Tobias Block
@tblock79
The other thing is that, unfortunately, the Siemens sequences change with every software version and are a "blackbox". So even if you have the raw data, you cannot say what the scanner was doing to acquire that data. We are currently working on a project to potentially solve that aspect (at least for custom-developed sequence only)
For our research work, we use our own in-house developed sequences and perform our reconstructions offline (using the ORT client + YarraServer) which is then sent to PACS
So for these scans, we then have the raw data and can run the reconstruction later.
BlakeGFitch
@BlakeGFitch
So, I think locally they make their own sequence for some stuff. The people doing work closer to the machine do their own sequences. For the "higher level" work, the neurosci folks are using a mix of siemens reconstruction and in-house reconstruction.
Which PACS do you use?
I'm sorry... have to head home to dinner. Very interested to hear more however. ttyl.
Tobias Block
@tblock79
So for that, the only option would be to develop own offline reconstruction modules. However, this will not exactly generate the same images that come from the scanner due to things like distortion correction, image normalization, image filtering etc which Siemens is not sharing
We are using Philips iSite, but the Yarra offline reconstruction should work with any PACS
BlakeGFitch
@BlakeGFitch
Just a quick update... it looks like we are cleared to try out RDS
Tobias Block
@tblock79
That's great! Hope it will go well
BlakeGFitch
@BlakeGFitch
I was wondering if you have an opinion about http://ismrmrd.github.io/
... the local opinion seems to be that we would lose information from the raw .dat header moving to that format. Have you considered using that stuff?
BlakeGFitch
@BlakeGFitch
I found what you wrote on ismrmrd in the FAQ.
pimpul
@pimpul
Hi, I am interested in the GRASP-VIBE sequence for Siemens. Who do I need to contact to set up a C2P?
Tobias Block
@tblock79
@BlakeGFitch It is possible to integrate conversion to ISMRMD as part of the Yarra reconstruction pipeline (e.g., as a preprocessing step). The Yarra framework is pretty much agnostic with regard to the format of the data. However, the scanner exports the data in the Siemens Twix format, so initially it will be in that format. I haven't worked with the ISMRMD format, but I heard that the converter does not work with the newer Siemens software versions. So this is something that I would check upfront (could be that it has been fixed by now).
@pimpul Please contact your local Siemens collaboration manager. It is required that your institution has a Master Research Agreement with Siemens, and you need to have an IDEA license. If both is the case, your collaboration manager should be able to prepare a C2P between Siemens and your institution. If that has been signed, we can send you the sequence/ICE binaries. Wish I could distribute it directly, but that's the policy required by Siemens, unfortunately.
pimpul
@pimpul
OK @tblock79 I will do that. We have a MRA and IDEA license
Tobias Block
@tblock79
Great! Let me know once that is in place and then I can send you a download link for the sequence
Jon Tamir
@jtamir
Hi all, thanks @tblock79 for the informational email about Yarra and pointing me to this community!
Jon Tamir
@jtamir
Hi, wondering if I can get some help exiting kiosk mode on the Vida
The instructions don't seem to work for me
Okay, the problem was my num-pad wasn't labeled and I got confused. I'm able to enter the login page, but I need to ask my site for the username/password
Tobias Block
@tblock79
I think Siemens uses the same password on every scanner worldwide. Will email it to you
hyunsoolee0503
@hyunsoolee0503
Hi, I've recently installed Yarra rds.exe on vida (XA11a) scanner and set the transfer time to "start up" since we turn off the scanner at night and reboot in the early morning everyday. It seemed that both "transfer now" and "start up automatic transfer" didn't work properly even if the yarra is manually turned on after rebooting the scanner.
  • The error messages are like: "Error: Could not access network drive path / Error: Check if network drive has been mounted. Or check configuration..
    I wrote all the login information in the configuration using "net use : ~", but only after I connected to network drive where the raw data should be stored with manual login process, "transfer now" worked.
    Can you suggest how I can handle this problem?
skwarfield
@skwarfield
When RDS takes the raw data from Twix and stores it on the local hard drive, what ensures that the local hard drive doesn't run out of disk space ?
Tobias Block
@tblock79
@skwarfield RDS calculates how much space is needed for the export and then checks if sufficient disk space is available (with a safety margin). If enough space is available, it will export all files to the drive and move them to the network share afterwards. If space is low, it will go into an "alternating" mode in which it exports one file at a time, moves it to the network share, and continues with the next file. If insufficient space is available, it will not export the files and show an error message. If Yarra LogServer has been configured to monitor the Yarra clients, Yarra LogServer will be notified about the problem (so that it can trigger an alert to Slack or so)
Tobias Block
@tblock79
@hyunsoolee0503 Sounds like there is a problem with the mapping of the network share. Check if the information entered on the Network page of the RDS configuration is correct (you could check if the information entered under Connect Cmd will connect the share successfully if executed on the command line). As an alternative, you could enter the command for restoring the network connection on the page "Startup Commands". These commands will be executed on the first start of RDS (like a batch file). We are using this mechanism to restore all network shares after reboot (we have "net use * /delete" as the initial command there to ensure that all prior network connections, which people might have setup accidentally, are cleared).
hyunsoolee0503
@hyunsoolee0503
@tblock79 Thanks to your help, I could find out that there was a mistake in the command line in the network page. Your advice really helped me a lot. Thank you:) Now the connection seems to work well and there is no error in the log. However, the rds.txt file in the connected network drive still didn't update the information after the installation day although the log in the Yarra program updates properly. Does this mean that there's still a connection problem? Should I remove the scanner folder (i.e., Vida175891) and try again?
Tobias Block
@tblock79
@hyunsoolee0503 Great to hear that it is connecting the drive now! Can you give some more information on which file RDS.txt you are referring to? I don't think that RDS is creating such a file
hyunsoolee0503
@hyunsoolee0503
image.png
@tblock79 This rds.txt file is located in the folder where the raw data is stored. This txt file includes the log information.
image.png
sbaete
@sbaete
@tblock79 Tobias, yarra is great! I can send dicoms to our PACS. I'm just having an issue making them show up in the right 'study' together with the images which were exported from the scanner. I think it has to do with some of the UIDs. Do you know if you can recover the StudyInstanceUID and/or StudyID from the dat-file?
hyunsoolee0503
@hyunsoolee0503
@tblock79
@tblock79 'rds.txt' --> 'rds.log'
Tobias Block
@tblock79
@hyunsoolee0503 Ah, I see. You mean the log file that is copied to the network share after the data transfer. Could you check if there are any errors or warnings listed in the main log file? It is located in the subfolder "log" of your Yarra installation, e.g. C:\yarra\log\rds.log? This log file is only copied to the network location if Yarra was able to run the connection command successfully, so maybe the log file on the C: drive contains additional information on what is going wrong here.

@sbaete Hi Steven, I don't think the UIDs have to match for that (I don't think we do that for the GRASP recons -- and they show up in the correct study on the PACS). Are you setting the Accession number (ACC) in the DICOM tags? This is necessary for the right assignment in PACS. Unfortunately, the ACC is not contained in the Twix files. However, if you configure your recon mode with

[ClientConfig]
RequiresACC=TRUE

then the Yarra client will ask for the ACC during submission and you can then read the ACC from the .task file (or by using the macro %vacc in the .mode file)

sbaete
@sbaete
@tblock79 Ah, ok, not sure what is going wrong then. I do set the right Accession number. I will have a closer look at the SetDCMTags-module to see how I can improve my dicom-tags. Thanks!
sbaete
@sbaete
@tblock79 Aha, now it works for a volunteer. I was trying it on a phantom, and I guess there might be other issues there. Thanks!
pimpul
@pimpul
@tblock79 a couple of weeks ago I asked about obtaining the GRASP-VIBE sequence. Siemens does not support a C2P for this sequence any longer. I've heard from Steven Sourbron that he uses the RAVE sequence for DCE. Can I obtain that sequence? Thanks in advance
Tobias Block
@tblock79
@pimpul Hi! The RAVE sequence is the sequence I was referring to. It's our own implementation of a radial stack-of-stars 3D GRE sequence (similar to the StarVIBE / GRASP-VIBE sequence) and we have provided it to various collaboration partners. However, Siemens requires a C2P agreement before we can share the sequence binaries. Could you reach out to your Siemens collaboration manager regarding the C2P? The C2P agreement has to be between Siemens and your institution. Once it has been signed, we can send you the sequence binaries. The deliverable would be the RAVE sequence and corresponding ICE program.
Dario Bosch
@dabosch
Hi,
I'm working together with @BlakeGFitch and have a question regarding RDS.
I noticed that RDS first exports the data files to a local folder (queue) before moving them to the specified network drive. Especially for large data files that overall process takes significantly longer than directly exporting to the network drive using twix. Is there a reason why no direct export is done? Are there plans to change that in the future, or add it as an option?
On our system the local hard drive seems to be slower than the network drive so I would prefer a direct export.
Tobias Block
@tblock79
@dabosch Hi Dario, The scanners cannot start new scans while data is being exported. Therefore, Yarra is first writing all data to the C: drive and then moving it to the network share (during which scanning can proceed). Also, this allows buffering data if the network share is temporarily unavailable. It would very much surprise me if the network share is faster than the local hard drive (I think generally, this is not the case). However, if you want to test it, you could a) install your Yarra client on the network share instead of the local drive (because it's buffering the files in the "queue" folder under the installation path) or b) modify the code accordingly and rebuilt the client