I am using fooof to analyze working memory task data, as mentioned in this article Parameterizing neural power spectra into periodic and aperiodic components. However, when I conducted PSD analysis on the trials correctly reflected by the subjects, I found that the time window for each trial was too short, resulting in a very short PSD window and very few segments. The PSD results obtained cannot be further analyzed. How can How can I do the same task data analysis as the article by using Brainstorm?
@Luc have you replicated such article within Brainstorm?
Another question is if I want to analyze the original psd and the psd after correcting the aperiodic activity, which part of the data should I use in the fooof structure exported to matlab?
For the peaks, only the center-frequency, amplitude and standard_deviation of each peak is saved in the FOOOF structures. A detailed description of the FOOOF data can be found in here:
https://neuroimage.usc.edu/brainstorm/Tutorials/Fooof#Accessing_FOOOF_model_parameters
If you want to obtain the frequency-resolved periodic activity, it is possible to extract it with as shown in here:
https://neuroimage.usc.edu/brainstorm/Tutorials/Fooof#Convert_FOOOF_model_parameters_to_regular_PSD_files
By doing this, a standard PSD file will be created with the requested FOOOF metric, this allows using all the visualizations available for PSD: 2D 3D sensor cap, 2D disk (topoplot), 2D Layout, ...
Also, Extracting the FOOOF measures, will allow using the 2D Layout plot for multiple metrics: e.g., Spectrum (original PSD) and Aperiodic
