[P069] Optimal B-values for diffusional kurtosis imaging of the liver and pancreas

Purpose Diffusional Kurtosis Imaging (DKI) is a new, rapidly developing magnetic resonance (MR) diffusion weighted imaging (DWI) technique. DKI requires acquisition with several b-values, including high b-values (>1000 s/mm 2 ). DKI takes into account the influence of tissue barriers (cell membranes, organelles) on diffusion, and is presumed to better reflect the actual water diffusion processes than standard DWI. However, several limitations hinder use of DKI in routine practice, including lack of standardization and long acquisition time (due to additional b-values). We aimed to optimize the DKI method to ensure reliable results with the shortest possible acquisition time. Methods In this prospective study 20 healthy volunteers (10 men, 10 women; age: 25–62 years; mean: 39) underwent DKI at 3.0 T Siemens Magnetom Skyra using 7 b-values (0; 200; 500; 750; 1000; 1500; 2000 s/mm 2 ). Region of interests (ROIs) were placed in liver (right lobe, left lobe) and pancreas (head, tail). DKI parameters (D app , K app ) for ROIs were calculated for 7 b-values using program written in Gnuplot, utilizing the nonlinear least-squares (NLLS) Marquardt–Levenberg algorithm. D app and K app quantitive maps were created using the in–house software based on ImageJ. All calculations were repeated for five subsets of data, with number of b-values reduced to 4 and 5. DKI parameters calculated for subsets were compared with parameters calculated for all 7 b-values (Wilcoxon signed rank test). Statistical analysis was performed in R; p  Results Correlation coefficient between DKI parameters calculated for 7 b-values and subsets ranged from 0.69 to 0.99. Coefficient of variance (CoV) of parameters calculated for group of volunteers varied from 11% to 33%, with lowest agreement for head of pancreas (D app ) and left lobe of liver (K app ), and was not affected by reduction of b-values number. Significant differences between DKI parameters calculated for 7 b-values and subsets were found for all subsets except one (0; 500; 1500; 2000 s/mm 2 ), for which no differences were observed. Conclusions DKI acquisition with only 4 b-values (0, 500, 1500, 2000 s/mm 2 ), compared to DKI acquisition utilizing 7 b-values, allowed for the reduction of acquisition time by 36%, without affecting calculated DKI parameters.