Functional connectivity (FC) estimates are sensitive to choice of estimation method (e.g., sliding windows, MGARCH, Wishart processes, hidden markov models). This project aims to design a range of benchmarks to determine what estimation method to use.

These benchmarks have been particularly developed to test the estimation methods included in the FCEst Python package. Results have been published in an article in Imaging Neuroscience (see CITATION.cff).

Many extensions of this project are possible, both in terms of adding more estimation methods and more benchmarks.

This is an open-source project and contributions are more than welcome. Please raise an issue here on Github or send me a message.

A curated list of relevant publications related to FC estimation benchmarking can be found on Semantic Scholar.