027108

A Library of Human Electrocorticographic Data and Analyses

Kai Miller

Introduction: Electrophysiological data from implanted electrodes in the human brain are rare, and therefore scientific access to it has remained somewhat exclusive.

Objective: In order to democratize the use of ECoG data to understand brain function, we have collected and standardized a large volume of recordings and the code to help understand them from a variety of behavioral experiments.

Methods: Here we present a freely-available curated library of implanted electrocorticographic (ECoG) data and analyses for 16 behavioral experiments, with 204 individual datasets from 34 patients recorded with the same amplifiers, at the same settings. For each dataset electrode positions have been carefully registered to brain anatomy. The accompanying MATLAB script library is tailored specifically to identify and extract high-yield features from the ECoG signal and is embedded in the library alongside the data.

Results: All data, anatomic locations, and analysis files (MATLAB code) are in a shared file structure at https://searchworks.stanford.edu/view/zk881ps0522. They may be accessed from any location with no registration necessary, and used without restriction.

Conclusions: This library represents the first standardized benchmark of ECoG data and code. With the existing analysis structure, a backbone is in place, where data are loaded, variables accessed, figures made, and plots cycled through – one may simply insert their code at the appropriate place within scripts and have results plotted on anatomy instantly. The library may be used as course material or serve as a starter package for researchers early in their career or for established groups, to modify the analyses and re-apply them in new settings.