Optimal HRF modeling for BOLD–neuronal coupling

Determine the most appropriate hemodynamic response function (HRF) model that accurately captures the coupling between observed blood-oxygen-level-dependent (BOLD) signals and underlying neuronal firing in functional MRI analyses, specifying the model class, parameterization, and estimation strategy that best represent this relationship across typical study designs.

Background

The paper describes afni_proc.py’s extensive support for hemodynamic response modeling via AFNI’s 3dDeconvolve, offering fixed-shape, variable-shape, and hybrid basis functions, as well as options for amplitude and duration modulation. Despite the tooling flexibility, the authors emphasize that the fundamental question of how to best model the link between measured BOLD signals and underlying neuronal activity remains unresolved.

This uncertainty affects task-based analyses, interpretation of effect estimates, and cross-study comparability, motivating continued methodological development and validation of HRF formulations and parameterizations within FMRI pipelines.