In vivo MR acoustic radiation force imaging in the porcine liver.

PURPOSE High intensity focused ultrasound (HIFU) in the abdomen can be sensitive to acoustic aberrations that can exist in the beam path of a single sonication. Having an accurate method to quickly visualize the transducer focus without damaging tissue could assist with executing the treatment plan accurately and predicting these changes and obstacles. By identifying these obstacles, MR acoustic radiation force imaging (MR-ARFI) provides a reliable method for visualizing the transducer focus quickly without damaging tissue and allows accurate execution of the treatment plan. METHODS MR-ARFI was used to view the HIFU focus, using a gated spin echo flyback readout-segmented echo-planar imaging sequence. HIFU spots in a phantom and in the livers of five live pigs under general anesthesia were created with a 550 kHz extracorporeal phased array transducer initially localized with a phase-dithered MR-tracking sequence to locate microcoils embedded in the transducer. MR-ARFI spots were visualized, observing the change of focal displacement and ease of steering. Finally, MR-ARFI was implemented as the principle liver HIFU calibration system, and MR-ARFI measurements of the focal location relative to the thermal ablation location in breath-hold and breathing experiments were performed. RESULTS Measuring focal displacement with MR-ARFI was achieved in the phantom and in vivo liver. In one in vivo experiment, where MR-ARFI images were acquired repeatedly at the same location with different powers, the displacement had a linear relationship with power [y = 0.04x + 0.83 μm (R(2) = 0.96)]. In another experiment, the displacement images depicted the electronic steering of the focus inside the liver. With the new calibration system, the target focal location before thermal ablation was successfully verified. The entire calibration protocol delivered 20.2 J of energy to the animal (compared to greater than 800 J for a test thermal ablation). ARFI displacement maps were compared with thermal ablations during seven breath-hold ablations. The error was 0.83 ± 0.38 mm in the S/I direction and 0.99 ± 0.45 mm in the L/R direction. For six spots in breathing ablations, the mean error in the nonrespiration direction was 1.02 ± 0.89 mm. CONCLUSIONS MR-ARFI has the potential to improve free-breathing plan execution accuracy compared to current calibration and acoustic beam adjustment practices. Gating the acquisition allows for visualization of the focal spot over the course of respiratory motion, while also being insensitive to motion effects that can complicate a thermal test spot. That MR-ARFI measures a mechanical property at the focus also makes it insensitive to high perfusion, of particular importance to highly perfused organs such as the liver.