[Eeglablist] ROI Connect Power Extraction

[Cedric Cannard]

Hi Edward,

Moving this back into EEGLAB list so it can potentially help others with the same question later.

- The documentation mentions that this variable is the sum of the PSDs for the relevant voxels for each ROI. Is this done because PSDs from individual voxels are noisy, and averaging the PSD across relevant voxels would not be recommended?

Yes reduce noise and reduce dimensionality of the data, you would end up with thousands of voxels, each representing a small region with time-varying activity. ROIconnect uses PCA (3 fixed components) to aggregate the activity per ROI, retaining the dominant patterns to reduce dimensionality, noise, and facilitate connectivity analysis.

See the paper for more details: https://urldefense.com/v3/__https://www.sciencedirect.com/science/article/pii/S1053811923003695?via*3Dihub__;JQ!!Mih3wA!CJdYecXqflnCN_nJ-oWXpyfMOmTPXujtpt2sDpTpKVFK0QZ3JaudBF2wYhvvMI4x7jQ5DZJc0DPZZ8PfCnHe8DmZOA$ 

Cedric Cannard

On Friday, October 25th, 2024 at 2:43 PM, Edward Lannon <elannon at stanford.edu> wrote:

> Thanks Cedric! You are right. EEG.roi.source_roi_power is probably the right variable since the dimensions fit the Frequency X ROI format. The documentation mentions that this variable is the sum of the PSDs for the relevant voxels for each ROI. Is this done because PSDs from individual voxels are noisy, and averaging the PSD across relevant voxels would not be recommended?
>
> Best,
>
> Edward
>
> ---------------------------------------------------------------
>
> From: Cedric Cannard <ccannard at protonmail.com>
> Sent: Friday, October 25, 2024 12:36 PM
> To: Edward Lannon <elannon at stanford.edu>
> Subject: Re: [Eeglablist] ROI Connect Power Extraction
>
> If I recall correctly, norm is normalized across all ROIs, and I would typically use EEG.roi.source_roi_power to look at specific regions or networks. But could be wrong. what size are the variables? that would easily confirm it.
>
> Cedric
>
> On Monday, October 21st, 2024 at 5:52 PM, Edward Lannon via eeglablist <eeglablist at sccn.ucsd.edu> wrote:
>
>> Hello,
>>
>> I am using the ROI connect plug in. If I wanted to get the PSD from the ROIs, would I use the EEG.roi.source_roi_power or the EEG.roi.source_roi_power_norm variable?
>>
>> Thanks for the help!
>>
>>
>>
>> _______________________________________________
>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
>> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu