[FieldTrip] Explanation about the Fieldtrip-SimBio pipeline.

[Schoffelen, J.M. (Jan Mathijs)]

Hi Yiannis,

We looked into this in more detail, and our conclusion is that the first row of the transfer matrix indeed is intended to be zero. That is, the low-level computations (in external/simbio) by construction compute the potentials referenced to the (arbitrary) first electrode in the electrode list. Thus the for-loop starting at ‘2’  in sb_transfer is intentional.

The aim of the FieldTrip code, is to return the EEG leadfields average referenced, unless explicitly defined by the user. Somehow this did not go entirely well in this case. We will fix this in the code. Thanks for reporting.

For now, in order to be able to use the computed leadfields in combination with the data, you need to ensure that the referencing scheme in the data is consistent with the referencing scheme for the leadfields. It would make sense to average reference both, or to reference the data against the first channel (as per the data.elec.label(1)).

Best wishes,
Jan-Mathijs


On 2 Nov 2021, at 13:35, Ioannis Zorzos via fieldtrip <fieldtrip at science.ru.nl<mailto:fieldtrip at science.ru.nl>> wrote:

Dear all,

Recently, I came across something peculiar regarding the Leadfield matrix and the transfer matrix of the conduction volume model, created after following the Fieldtrip-SimBio pipeline.

The Leadfield matrix (from now on called LF) that is produced (which is a field inside the leadfield struct created from the ft_prepare_leadfield function) is a {1 xN_s} cell array, where N_s is the number of sourcepoints specified during the creation of sourcemodel. Assuming that all the source points are “inside” the brain (i.e. the “inside” field of the leadfield struct would be a column  {N_s x 1} full of ones), then LF contains N_s  cells, each one being of size {N_ch × 3 double}, where N_ch is the number of channels/electrodes used.

I noticed that the first row of each of these cells from the LF (i.e. for the first channel/electrode) for each of the source points is zero. This is obviously due to the transfer matrix, of which the first row is also comprised of zeros, and is created with the sb_transfer.m function. In this function, the creation of the transfer matrix clearly starts from the second in order channel/electrode (the for loop starts from the second channel).

Is there a reason for this? Shouldn’t it start from the first channel?
Thank you in advance.

Sincerely,
Zorzos Yiannis

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://urldefense.com/v3/__https://doi.org/10.1371/journal.pcbi.1002202__;!!HJOPV4FYYWzcc1jazlU!ul2oMKwSlIeyLs3KdD7YtVe6QU587xFSOgvHNjrkMWanPIKg5VASFzKyqL71LWSalH3-DZiFjGb_0lw$

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20211110/a94c68d7/attachment.htm>