About PTE values at sensor level

1

Good nights!

I was comparing PTE values from one EEG file with and without normalization, but values dont seem to keep proportional results
I mean:

In no normalized version I got values between 0.19 and 0.27, the highest value was 0.2711 for F8-T3 pair,

in normalized version values I got are around 0.01 and 0.04 (some with minus signus), however, for F8-T3 pair the value I got was: 0.010 wich is one of the smallest I got.

Normalized option gave me a value of 0.044 for P4-Fp1, wich is the highest value in the connectivity matrix . However with non normalized option values for that pair was 0.26 wich is higher than some pairs in the matrix, but not the highest one

I dont know if normalization process changes proportions,

wich value is the correct one?

Thanks in advance and sorry for language issues

3

Our lab usually relies on normalized PTE values. I am not sure about the reason for this apparent discrepancies. I encourage you to consult the method's seminal paper:
https://www.sciencedirect.com/science/article/abs/pii/S1053811913009191

6

I think I finally understood this issue! (maybe):
the difference in the PTE values (between normalized and non-normalized) is just because non normalized PTE values shows the connectivity direction (given by the highest pair value juts as Granger causality) and normalized PTE does not!

Am I right?!!

PTE is actually an asymmetrical measure (directed), so maybe it could be color coded (in graph) just like Granger causality in BS,
What do you think?
:slight_smile:

Thanks in advance!

7

@Sylvain @hossein27en ?