Modeling dynamic inflow effects in fMRI to quantify cerebrospinal fluid flow.

Area of Science:

• Neuroimaging
• Biophysics
• Medical Imaging Analysis

Background:

• Cerebrospinal fluid (CSF) flow is critical for brain health and requires quantitative imaging techniques.
• Flow-sensitive functional MRI (fMRI) offers high sensitivity and temporal resolution for CSF dynamics but lacks quantitative measurement.
• Existing fMRI signals for CSF flow are not directly quantitative, limiting their interpretability.

Purpose of the Study:

• To develop a dynamic model for simulating fMRI inflow signals based on time-varying flow velocities.
• To validate the model using human and phantom data for accurate interpretation of fMRI inflow signals.
• To create a physics-based deep learning framework for direct velocity estimation from fMRI inflow data.

Main Methods:

• Development of a dynamic model to simulate fMRI inflow signals.
• Validation of the dynamic model using both human and phantom datasets.
• Implementation of a physics-based deep learning framework to invert the dynamic model for velocity estimation.

Main Results:

• The dynamic model accurately simulates fMRI inflow signals, providing insights into signal interpretation.
• The physics-based deep learning framework successfully enables direct estimation of CSF flow velocity from fMRI data.
• The developed methods allow for quantitative analysis of CSF flow dynamics using fMRI.

Conclusions:

• This work introduces a novel approach to quantitatively analyze cerebrospinal fluid flow using fMRI.
• The developed dynamic model and deep learning framework enhance the physical interpretability of fMRI flow signals.
• Researchers can now leverage the sensitivity, temporal resolution, and availability of fMRI for quantitative flow measurements.