Click here to


Are you sure ?

Yes, do it No, cancel

Intermachine Reproducibility of R2* Measurements Using Least-Squares Fitting

A Headley*, J Grice, D Pickens, Vanderbilt University, Nashville, TN


(Saturday, 4/4/2020) 10:30 AM - 12:30 PM [Mountain Time (GMT-6)]

Purpose: Constructing a phantom with low-cost, readily available materials that can be used to quantify transverse relaxation rate (R2*=1/T2*) and using the phantom to assess reproducibility of R2* quantification across four 1.5T MRI systems. This validation allows comparison of R2*-derived quantities, like iron concentration and fat fraction [1] across 1.5T MRI systems.

Methods: The phantom contains 4 vials with R2* values that cover a clinically-relevant range (100 to 420 Hz) [2]. Iron content was achieved using ferric chloride solutions contained in glass vials, each affixed in a 3D-printed polylactide structure, surrounded by distilled water, all housed in a sealed acrylic cylinder. Images were acquired on three scanners using the same mFFE pulse sequence and a least-squares curve fitting technique was used to calculate R2*. On the fourth scanner, images were acquired with different TEs, TR, and flip angle, using the same calculation method for R2*.

Results: On average, the intermachine variability in measured R2* values for the three scanners using the same acquisition parameters was 2.4% in the lowest-iron-concentration vial, 1.1% for the next highest iron content, 1.7% for the next highest, and 1.4% for the highest. The variation between all four scanners was less than 9% in all vials.

Conclusion: Producing accurate concentrations of ferric chloride was not possible, but the phantom proves useful for intermachine comparison. The phantom was used to determine that R2* quantification is highly reproducible across three 1.5T MRI systems using the same acquisition parameters. However, larger variability was introduced on the fourth scanner when altering acquisition parameters. Still, R2* measurements can be used to yield liver iron estimates that are more reproducible than biopsy overall [3]. Nevertheless, altering acquisition parameters increases variability, so an intermachine reproducibility evaluation should be conducted before multiple scanners are used to assess patient R2* for clinical parameter determination.

Contact Email