Project code for extracting neural population timescale from field potential data (LFP, ECoG, etc).

Paper is now published in eLife (here.)

Gao, R., van den Brink, R. L., Pfeffer, T., & Voytek, B. (2020). Neuronal timescales are
functionally dynamic and shaped by cortical microarchitecture. eLife, 9, e61277.

Neuronal timescales are functionally dynamic and shaped by cortical microarchitecture.

In this project, we developed a method for measuring neuronal timescales from neural field potential data via spectral parameterization, and apply it to invasive ECoG data from humans and macaques. We find a gradient of neuronal timescales that increase from sensory/motor towards association brain regions, and further combine several other brain-wide structural, gene expression, and behavioral datasets to dissect the physiological factors that underly variations in timescale across the brain, as well as its change during behavior and aging.

This project uses several open datasets, thanks to the generosity and foresight of those that compile and share their data. See Table 1 in the paper for a list of all datasets used.

./echo_utils.py contains all the python helper functions used for subsequent analyses and visualizations.

./scripts/ contains analysis scripts that compute and parameterize the PSDs in each ECoG database.

./data/ contains intermediate data tables and diagnostic plots.

./notebook/ contains Jupyter notebook that explains the project and paper in its entirety, and produces the figures seen in the publication. See Table 2 in the paper for the notebook-figure correspondence.

Surface projection of T1w/T2w and gene expression data is done using Rudy's repository here.