Phantom evaluations of low frequency MR elastography.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Rheology

Background:

• Non-invasive estimation of brain mechanical properties is crucial for understanding neurological disorders.
• Existing MR elastography (MRE) methods often require external mechanical drivers, causing discomfort and limiting applications.
• Intrinsic activation MR elastography (IA-MRE) leverages inherent low-frequency brain motion, eliminating external hardware.

Purpose of the Study:

• To evaluate the performance of low-frequency (1 Hz) IA-MRE in phantoms.
• To assess the efficacy of non-linear inversion (NLI) using viscoelastic and poroelastic models for IA-MRE data.
• To establish baseline performance metrics for future brain stiffness studies.

Main Methods:

• Experiments were conducted using four gelatin phantoms.
• Low-frequency (1 Hz) MR elastography was employed.
• Non-linear inversion of viscoelastic and poroelastic models was used to analyze displacement measurements.

Main Results:

• Stiffness resolution of 6 mm (contrast ratio 4.21) and 9 mm (contrast ratio 1.77) was achieved.
• Stiffness edge response distance was measured at 9 mm.
• High intraclass correlation (0.93) was found between mechanical testing and poroelastic estimates, despite model-data mismatch.
• Poroelastic MRE outperformed viscoelastic MRE in inclusion detection and consistency with mechanical testing.

Conclusions:

• Low-frequency IA-MRE is a valid technique for assessing spatially resolved brain stiffness changes non-invasively.
• Poroelastic modeling provides a more robust framework than viscoelastic modeling for low-frequency IA-MRE.
• This study provides essential performance metrics for future in-vivo IA-MRE applications in neuroscience.