Confusing values of normalized PTE (?)

Raymundo.Cassani · January 18, 2022, 7:55pm

Zero-centering of normalized PTE values has been removed from Brainstorm.

Remove centering around 0 in normalized PTE results

brainstorm-tools:master ← rcassani:pte-normalization

opened 04:01PM - 17 Jan 22 UTC

Removes the centering around 0 for normalized PTE (dPTE) values, as it can lead …to wrong interpretations, as reported in: https://neuroimage.usc.edu/forums/t/confusing-values-of-normalized-pte/32069 Normalized values are computed as in [Hillebrand et al., 2016](https://www.pnas.org/content/113/14/3867). For example, normalized PTE values (before centering around 0) of dPTExy = 0.9 and dPTEyx = 0.1 indicate that information goes preferentially from x to y. Removing 0.5 on both metrics results in dPTExy = 0.4 and dPTEyx = -0.4, while the numeric difference between the metrics is kept, the magnitude of the metric meaning does not look as clear.