A dynamical model of the laminar BOLD response

Area of Science:

• Neuroimaging
• Biophysics
• Magnetic Resonance Imaging

Background:

• High-resolution functional magnetic resonance imaging (fMRI) using the blood oxygenation dependent level-dependent (BOLD) signal is crucial for non-invasively studying brain activity.
• The BOLD signal's hemodynamic origin causes temporal and spatial discrepancies with underlying neuronal activity, particularly the laminar BOLD response (LBR).
• The LBR is often blurred towards the cortical surface due to signal propagation in intracortical ascending veins.

Purpose of the Study:

• To present a novel cortical depth-dependent model of the BOLD response.
• To explicitly account for the influence of ascending and pial veins on cortical BOLD signals.
• To provide a framework for interpreting high-resolution fMRI data and estimating laminar neuronal activity.

Main Methods:

• Developed a new BOLD response model based on the principle of mass conservation.
• Incorporated the effect of venous blood flow on cortical depth profiles of the BOLD signal.
• Modeled BOLD signal dynamics as a function of physiological parameters and neuronal activity distributions.

Main Results:

• The model explains that the surface-ward increase in LBR is primarily due to increased baseline blood volume.
• Local maxima ('bumps') in LBR are attributed to spatially inhomogeneous neuronal activity, not local blood volume.
• Gradient-echo BOLD signal point-spread functions and inter-laminar delays are influenced by physiological parameters and neuronal activity, mediated by ascending veins.

Conclusions:

• The developed model offers a biophysical framework for interpreting experimental observations in high-resolution fMRI.
• It highlights the role of venous vasculature in shaping the laminar BOLD response.
• Future applications include deconvolution of hemodynamic bias to estimate laminar neuronal activity.