Unraveling Hidden Patterns of Brain Activity: A Journey Through Hemodynamic Deconvolution in Functional MRI

Abstract

Functional magnetic resonance imaging data analysis is often directed to identify and disentangle the neural processes that occur in different brain regions during task or at rest, and employs the blood oxygenation-level dependent (BOLD) signal of fMRI as a proxy for neuronal activity mediated through neurovascular coupling. The goal of this thesis is to enhance and expand techniques for identifying and analyzing individual trial event-related BOLD responses based on the Paradigm Free Mapping (PFM) algorithm, which utilizes a linear hemodynamic response model and relies on regularized least squares estimators to deconvolve the neuronal-related signal that drives the BOLD effect. Notably, these techniques estimate neuronal-related activity without relying on prior paradigm information.