Hi,
I am trying to plot the position information of the head surface and the channels using Polhemus patriot. I followed the tutorial, but the digitization is inaccurate. When I fix the source and reference and measure the position of the nasion multiple times with the stylus, there is a difference of several centimeters even though I am digitizing the same location.
On PiMgr, patriot is working correctly. Is there anything else I should check?
@mperron8, any suggestions and tips on using the Polhemus Patriot?
Just wanted to share that I’ve been experiencing the same problem with the newest version of Digitizer. However, the issue doesn’t seem to be present in the legacy version. I’ve been using that instead, and the localization accuracy has been noticeably better.
Hope that helps!
FYI @Marc.Lalancette
Even the legacy version doesn't seem to work. Even if I measure at the location of nasion three times, the same coordinates did not appear. The same coordinates are output each time. It seems that the information from patriot is not being received. I have followed the tutorial to set up the COM port, etc. If you have any advice, please let me know.
If one lab is able to get proper digitization with the legacy version, but not the other lab, it seems to indicate some configuration or hardware issue.
If none are getting proper digitization with the recent version, then there's something wrong in our code as well. Unfortunately, unless we're able to get our hands on a Patriot, it's much harder and time consuming to investigate, if we can even find anything. I'll ask around to see if an EEG lab here might have a Patriot, but it's a long shot.
Thank you very much for the valuable information.
I would like to confirm if my understanding is correct:
I plan to use individual MRI data to estimate the source of activities with the highest accuracy.
In this context, is Fastrak currently the most recommended option?
We are also prepared to use a 3D scanner if that would provide better accuracy.
Could you please advise us on your current recommendation?
Hello,
We cannot confirm your point 2, but one user, who commented above, indicated they could only get it working properly with the legacy version. I don't think it's that the new panel gives "worse accuracy", but that it's not communicating properly with the Patriot, i.e. it's not working correctly.
We will eventually try to get the new version working well for Patriot, but this may take some time.
We've been using Fastrak here without issue, but of course the accuracy depends on your physical setup (e.g. nearby metal) and how you use the system (e.g. how the reference receiver is attached to the head), not just the hardware or Brainstorm. That's one of the reasons we updated the panel, to be able to verify the fiducials at the end to make sure nothing moved and the data is good. And of course this digitization only helps with one step of the coregistration, which will also depend on other factors like head motion.
I don't have experience with other systems so I don't feel I can recommend a specific system for "better accuracy". But I believe our collegues that implemented the Revopoint option are happy with it and perhaps they can comment.
Cheers
Hi there,
Thanks @Marc.Lalancette
From our experience at [@USC], we saved a significant amount of time during data acquisition. Scanning typically took no more than 5 minutes, and participants were not fatigued by the issues commonly associated with traditional digitizers.
Once the data is recorded, the electrodes localization process is completed semi-automatically using the 3D scans within Brainstorm.
You can refer to this tutorial for more details:
We also received positive feedback from UTH, which has been using this technique for several years now.
That said, we haven't tested the method across all EEG caps and models. Please refer to the tutorials for more details.
Hello,
@Marc.Lalancette
Thank you for your response.
We have been mindful of the receiver's position and nearby metal objects; however, the results have not been satisfactory.
It is possible that the PC we are using for measurement are contributing to the issue.
We sincerely appreciate your attentive support.
@tmedani
We will also try measuring with 3D scanning. Thank you for your suggestion.