This project organizes efforts to compute the differences in power spectral density (PSD) slope between aMCIs, SAs, and their controls, for both task-related (Go-NoGo, referred to as gng) and resting-state data (referred to as rs). PSD slope has previously been identified as a proxy for the measurement of neural noise by Voytek and colleagues. The slope of PSDs are computed at both the sensor and source level (source modeling is done through BESA).
For instructions on running a specific analysis, see docs.
The project is organized into a few different folders:
- data organizes project data into four directories:
- auxilliary: Contains csv files detailing participant behavioral and non-EEG data.
- gng: Contains EEG data for the Go-NoGo task.
- rs: Contains resting-state EEG data. These are further subdivided into 20-second segments and full recordings. For the full recordings, we have the original 66-channel recordings, as well as BESA-exported source models.
- runs: Separate, time-stamped analyses which result from running analyses located in
srcare stored here.
- docs contains the steps and protocols for each analysis.
- figures contains figures produced.
- results contains final figures and results.
- src organizes project analysis and preprocessing files into two main directories:
- gng: Contains source files for the Go-NoGo task.
- rs: Contains source files for the resting-state data, further subdivided into 20-second analyses and full recording analyses.
Instructions on estimating neural noise on a set during resting state are located in the docs:
docs/Resting state pipelind.md: For estimating neural noise on resting state data.docs/Go-NoGo pipeline.md: For estimating neural noise on Go-NoGo data.