(Sunday, 7/14/2019)
Room: ePoster Forums
Purpose: To investigate the intra-individual variability of hepatic diffusion parameters in intravoxel incoherent motion (IVIM) analysis, and determine correlations with proton magnetic resonance spectroscopy (¹H-MRS) measurements of steatosis.
Methods: Twenty-two patients, 9 with a liver fat fraction (FF) > 5%, were included, with a 3 Tesla abdominal magnetic resonance imaging examination, with breath-hold diffusion-weighted sequence (multi-band factor=3, and SPAIR; 8 b-values: 0, 10, 20, 60, 150, 250, 500, and 800 s/mm²) and ¹H-MRS (volume=(20 mm)³, TR=2000 ms, TE=20-35 ms, TM=10 ms, BW=1250 kHz, flip-angle=90°) to quantify hepatic steatosis. A test–retest experiment (n=3) was used to assess reproducibility and repeatability. Diffusion coefficients in the left and right hepatic lobes were compared (n=6). T-tests were performed to determine differences in ROI placement with significance at p<0.05. The fraction of fast-diffusion, fast-diffusion, and slow-diffusion (f, Df, Ds) were calculated from the IVIM images using nonlinear least-squares fitting. T₂ corrected FF was fitted using AMARES in jMRUI. The relationship between diffusion parameters and FF was evaluated by Spearman correlation and regression in R.
Results: Analysis of retest experiments in the right lobe yielded CoV=24.4,33.1,68.9% for f, Df, Ds respectively and the variability between lobes CoV=21.6,46.5,40.2%. For the 6 subject comparison, the right and left average CoV=50.0,50.1,67.2%. The f between the right and left lobes showed significant difference, p=0.02. The f and FF are correlated (rho=0.492,p=0.023). Df and Ds were not significantly correlated with FF (rho=-0.134,p=0.56;rho=-0.162,p=0.48 respectively). Linear regression indicated a 0.33 proportionality with FF, p=0.006. A two-sided t-test showed significant difference in the f between normals and steatosis p=0.02.
Conclusion: This data illustrates the potential for IVIM imaging to assess hepatic steatosis. Variability in the measurements appears systematic when compared to the retests, this is likely due to patients unable to sustain the breath-hold. Artifact mitigation and correction need to be considered further.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
