7 Tesla MRI Liver Fat Quantification in Mice: Data Quality Assessment

Area of Science:

• Biomedical Imaging
• Magnetic Resonance Technology
• Quantitative MRI

Background:

• Proton Density Fat Fraction (PDFF) is a key MRI/MRS metric for quantifying fat content.
• Robustness of PDFF measurements is crucial for accurate clinical and research applications.
• Current methods may lack detailed spatial error assessment.

Purpose of the Study:

• To evaluate the robustness of PDFF data derived from MRI/MRS.
• To implement a spatially resolved error estimation technique for PDFF analysis.
• To validate the error estimation method using both ex vivo and in vivo mouse liver data.

Main Methods:

• Acquisition of in vivo and ex vivo MRI/MRS data on a 7T small animal scanner.
• Calculation of PDFF maps using a magnitude-based approach with 24 echo times.
• Pixel-wise error estimation performed via propagation of uncertainty.

Main Results:

• Error maps successfully identified measurement errors as a cause for unexpected PDFF variations in explanted liver.
• In vivo error maps raised concerns about PDFF data quality, leading to identification of motion and bladder filling as error sources.
• The method demonstrated utility in distinguishing true biological variations from measurement artifacts.

Conclusions:

• Combining pixel-wise PDFF data with corresponding error maps enhances the specificity and spatial resolution of reliability evaluation.
• This approach provides a more reliable assessment of PDFF value accuracy.
• Spatially resolved error estimation is vital for robust quantitative MRI/MRS analysis.