Evaluating Measurement Stability in Glioblastomas Using Magnetic Resonance Elastography: Repeatability and...

Area of Science:

• Neuroimaging
• Biophysics
• Quantitative Biomarkers

Background:

• Magnetic resonance elastography (MRE) shows promise for intracranial tumor characterization, including glioblastoma.
• Evaluating measurement consistency and interobserver agreement is crucial for MRE's clinical adoption as a quantitative biomarker.

Purpose of the Study:

• To assess the repeatability and interobserver agreement of absolute and normalized shear modulus (|G*|), storage modulus (G'), and loss modulus (G″) in glioblastoma.

Main Methods:

• Prospective study involving 13 adults with histopathologically confirmed glioblastoma using 3T MRI.
• Two same-session MRE acquisitions were performed, with tumor segmentation by two independent observers.
• Repeatability assessed via repeatability coefficient and coefficient of variation; interobserver agreement via intraclass correlation coefficient.

Main Results:

• Absolute tumor |G*| and G' demonstrated high repeatability (coefficients of variation 1.42%-2.08%) and excellent interobserver agreement (ICC 0.97-0.95).
• Normalized measurements showed significantly lower repeatability (coefficients of variation 4.21%-5.45%) compared to absolute values.
• Tumor G″ exhibited higher variability, with no significant difference between absolute and normalized measurements.

Conclusions:

• Absolute tumor |G*| and G' are the most repeatable and reliable metrics for glioblastoma characterization using MRE.
• Normalization to normal-appearing white matter may decrease measurement repeatability due to white matter variability.
• The study's small sample size (n=13) warrants caution regarding generalizability.