Determining Excitatory and Inhibitory Neuronal Activity from Multimodal fMRI Data Using a Generative Hemodynamic...

Area of Science:

• Neuroscience
• Physiology
• Biophysics

Background:

• The blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) signal indirectly measures neuronal activity.
• BOLD signal transients can reflect neuronal activity or vascular changes like cerebral blood flow (CBF) and venous cerebral blood volume (venous CBV).
• Distinguishing neuronal from vascular contributions to the BOLD signal is crucial for accurate interpretation.

Purpose of the Study:

• To introduce and validate a novel generative hemodynamic model, P-DCM, within the dynamic causal modeling framework.
• To demonstrate the model's ability to jointly explain dynamic relationships between neuronal and hemodynamic variables using multi-modal data.
• To assess the model's capability in dissociating neuronal from vascular transients and inferring neuronal activity from BOLD signals.

Main Methods:

• Proposed a novel generative hemodynamic model (P-DCM) inspired by physiological observations.
• Utilized three distinct multi-modal experimental datasets from human, cat, and monkey primary visual areas.
• Fit the P-DCM to datasets measuring BOLD, CBF, total CBV, and neuronal activity.

Main Results:

• The P-DCM accurately modeled neuronal and vascular transients and effective connectivity.
• Demonstrated that BOLD signal dynamics do not always unambiguously indicate neuronal activity dynamics.
• Successfully dissociated neuronal from vascular transients and deduced neuronal activity time-courses from BOLD and multi-modal data.

Conclusions:

• The generative model, P-DCM, effectively accounts for the dynamics of physiological mechanisms underlying the BOLD response.
• The model allows for the separation of neuronal and vascular contributions to the BOLD signal.
• P-DCM enables the inference of excitatory and inhibitory neuronal activity from BOLD data alone and multi-modal datasets.