Import the fiducial file

Hellow, Thanks for support to BST users.

I would like to know about how to import fiducial file marked by brainsight, and will match 3 points (NAS/LPA/RPA). I read Brainstorm tutorial, then we can match head localization between MRI and MEG (CTF) using Polhemus device when put *.pos file on the MEG data set. BrainSight doesn't export *.pos (it can export *.txt, *slp/.hsp, *sfp, *.mat). We cannot import these data from degitized head points>add point in Brainstorm.
Thank you.

Hi Nojima,

While the tutorial happens to use Polhemus, Brainstorm is not limited to .pos files when importing head points. We support a variety of formats, including .sfp that you listed. To import your headpoints, right click on your channel file and select Digitized head points -> Add points… Then, in the selection box for “Files of Type”, you can see all the formats we support. For example, “EEG: EGI (*.sfp)”.

I hope this helps!


Hi Martin

I have exported MRI coordinates in a *.txt file from Brainsight. Based upon the tutorials, it appears we might manually enter the coil and fiducial locations in MRI coordinates (NIFTI (MRI) coordinates), we cannot read text data (select Digitized head points -> Add points -> select *.txt. -> it shows "No file could be read]"). do we need to write a plug-in scritp for Brainstorm?
Although we have recognized that it is a better way to digitize 100-200 additional points for the head shape of the subject, we will use the NAS/LPA/RPA points of the electrodes file to try matching the electrodes positions on the MRI.
If we can use *.txt file, can you give me information how to enter manually these data.



Hi Ippei

Could you give us some more information about what your are trying to do?
What kind of recordings are these? TMS+EEG?
How do you get this .txt? what does it contain? please share an example file.
Do you have the individual MRI scans of your subjects? Have you processed them with FreeSurfer or any other program? Have you imported them in Brainstorm?

Including a few screen captures showing the files in your database explorer may help us understand better.


Hi Francois

To analyze the MEG data, we got the txt file of fiducial locations by using the brainsight.

We have the individual MRI scans of our subjects, and have processed them with Freesurfer, then we imported them in brainstorm.

Any help is appreciated!



Hi Francois

I could not attach our dataset on this forum, so I have uploaded the example text file on web last time. Please let me know if you need other information about our issues.


Hi Ippei,

Brainstorm cannot read the positions from this brainsight file at the moment, but this is something relatively easy to add if needed.

However, I'm not sure I understand what you expect to do in Brainstorm with this file. The fiducials NAS/LPA/RPA are used only to get a basic approximation of the registration of the MRI and the sensors. Which means you need to have these fiducials marked in the Brainstorm MRI Viewer + these fiducials available in the same file as the sensor positions. Importing fiducials positions alone is useless...

Addionally, we recommend you don't limit your MRI-MEG registration only to 3 points, but use hundreds of additional head points instead:

What is your MEG system?
And what are you using Brainsight for?


Hi Francois

Our MEG system is CTF.
The position of 3 points (NAS/LDA/RDA) are not saved in the MEG datasets because we unfortunately don't have the original CTF software. Although I completely recognize that using hundreds of head points is better way to fit between MRI and MEG data, we will try to align 3 points on the MRI by these fiducial points using coodinate data (*.txt) of the Brainsight instead of a Polhemus FastTrak.

However, fiducial points from the Brainsight are not matched on MRI view (see the attached file). When we import the coodinate data into dataset, do we have to adjust the data from for brainsight to for brainstorm?

Brainsight coodinate
NAC 3.751 64.411 26.11
RAC 74.85 -4.06 -32.303
LAC -65.353 -4.604 -38.134

Thanks in advance


The goal of these 3 coordinates are to localize the head while it is inside the helmet during the recording. As François mentioned having them measured outside the MEG is useless unless it's to align other points (head points or EEG electrodes for example), but you would still need the MEG positions to do the alignment.

If you don't have the CTF software, do you power the fiducials at all? If so, then you could localize them yourself with magnetic dipole fits at the proper frequencies, assuming the sampling rate is high enough and the head coil frequencies are not being filtered out.

I'm afraid without localization of the head coils, you should probably limit yourselves to sensor-level analysis, in which case you don't have to worry about aligning to the MRI.