[FieldTrip] Value range of power from source analysis (eLORETA)

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

Hi Ina,

For distributed source reconstruction techniques (e.g. e/sLORETA MNE etc) the magnitude of the values you get scale with the number of grid points in your sourcemodel. The larger the number, the smaller the values. This is because of the underlying model, where the sensor data (with fixed magnitude) is explained as a linear combination of the underlying sources. The more sources, the lower the contribution of the individual sources.

I hope that this helps.

Jan-Mathijs


On 27 Jun 2022, at 23:18, Ioana Susnoschi-Luca (PGR) via fieldtrip <fieldtrip at science.ru.nl<mailto:fieldtrip at science.ru.nl>> wrote:

Hi everyone,

I was comparing my source analysis results for a source and leadfield model with resolution 1mm and one with resolution = 1cm and have noticed big differences in source-level power, as estimated with eLORETA. I am using the standard MRI from the Fieldtrip library.
In the case of res = 1mm, source power ranges between 1-15, while for res = 1cm, power values are much lower, 0.01-0.5. The sensor-level power (from ft_freqanalysis) ranges between 0.01-3.

When changing the model, I specified cfg.unit = 'cm' for the following functions: ft_prepare_headmodel, ft_prepare_sourcemodel, ft_prepare_leadfield. Is it possible I double-scaled somewhere?

Any indication of why this is the case or advice is very much appreciated!

Thanks,
Ina
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://urldefense.com/v3/__https://doi.org/10.1371/journal.pcbi.1002202__;!!HJOPV4FYYWzcc1jazlU!-Vq3gvZmWSbgq9xWzNAwNQWrhGiGGZMyAa0USzFi9egdJfb23_bWYCo5zOKMLqWlHUk4AjIE4ZCSMI6a9b_wnItZx20HfMmNWJdx9A$

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20220628/e74f7de9/attachment.htm>