Methods Series

Bi-Cross-Validation: A Data-Driven Method to Evaluate Dynamic Functional Connectivity Models in fMRI

Presented by Yiming Wei

Abstract: Functional connectivity (FC) quantifies interactions between brain regions. Dynamic functional connectivity (dFC), which captures temporal variations in these interactions during resting-state fMRI, has gained much attention recently. However, evaluating dFC models against each other and selecting optimal configurations remains challenging. In this talk, I will introduce bi-cross-validation (BCV), a data-driven approach designed to tune hyperparameters within models and compare performance across different dFC models. We have evaluated bi-cross-validation using large-scale datasets such as the Human Connectome Project (HCP) and UK Biobank (UKB), demonstrating its potential for robust model selection. We also encourage you to apply and compare these models on your own study-specific datasets using the osl-dynamics toolbox.