[Diffusion-weighted echo-planar sequences for the evaluation of the upper abdomen: technique optimization].

PURPOSE

To optimize the technique for the evaluation of molecular diffusion in the abdomen.

MATERIAL AND METHODS

Fifteen healthy volunteers, 6 males and 9 females, ranging in age between 24 and 31 years underwent an MRI evaluation of the upper abdomen, using a superconductive 1.5T magnet (maximum gradient strength, 25 mT/m; minimum rise time 600 ms), equipped with phased array abdominal multicoil. Diffusion study was performed with a single-shot Inversion Recovery Spin-Echo Echo-planar sequence (IR-SE-EPI) with the following parameters: TR = infinite; TE=101 ms; matrix 128 yen 128; receiver bandwidth 2080 Hz/pixel; slices: n.20; slice thickness: 8 mm; acquisition time: 5.41 s. For diffusion weighting the following b values were employed: b=30 mm/s2, b=300 mm/s2 e b=500 mm/s2. Both qualitative and quantitative (calculation of linear regression analysis and of apparent diffusion coefficient) image analysis was performed.

RESULTS

Image quality was graded as diagnostic in all the cases. Image quality decreased with the increase of b values: at low b values, the anatomy of upper abdominal organs was easily recognized, whereas, at high b values, the same organs could not be adequately assessed unless the images were compared with those obtained with low b values. Magnetic susceptibility artifacts were observed in all the cases; no significant chemical-shift artifacts were observed as the fat saturation pre-pulse was employed. Quantitative analysis demonstrated an apparent diffusion coefficient of 1.58 s/mm2 for the liver, 1.61 s/mm2 for the spleen and 5,14 s/mm2 for the gallbladder. A statistically significant difference (p<0.001) was observed between parenchymatous organs (liver and spleen) and gallbladder, presenting as a stationary fluid.

CONCLUSIONS

Diffusion-weighted MR sequences may be implemented for abdominal studies, but the optimization of same parameters is slightly different compared with neuroradiologic applications. The potential applications are interesting above all as regards the characterization of focal liver lesions. Further developments are awaited in both sequence optimization (greater stability and lower sensitivity to magnetic susceptibility artifacts) and data analysis, with more complex algorithms able to better quantify the real diffusion coefficient.