Adaptive focusing of transcranial therapeutic ultrasound using MR Acoustic Radiation Force Imaging in a clinical environment

Background: In order to focus ultrasound beams through aberrating layers such as fat or bones, adaptive focusing techniques have been proposed to improve the focusing, mostly based on the backscattered echoes. We recently proposed an energy-based technique with the sole requirement being knowledge of the acoustic intensity at the desired focus. Here, Magnetic Resonance-Acoustic Radiation Force Imaging (MR-ARFI) is used to map the displacement induced by the radiation force of a focused ultrasound beam. As the maximum displacement is obtained with the best corrected beam, such a measurement can lead to aberration correction. Material and methods: Proof of concept experiments were previously shown in a small animal MR at 7 T using a 64-elements linear phased array operating at 6 MHz. Optimal refocusing was then obtained through numerical and physical aberrating layers. This work is extended here in a clinical Philips 1.5 T Achieva scanner. The HIFU beam is generated using a 512 elements US phased array (SuperSonic Imagine, France) dedicated to transcranial human experiments and operating at 1 MHz. Experiments are conducted in phantom gels and ex vivo brain tissues through numerical phase aberrators. A motion-sensitized spin echo sequence (TE = 70 ms, TR = 1200 ms, spatial resolution is 2×2×7 mm 3 ) is implemented to measure displacements induced by the acoustic radiation force of transmitted beams. Results: MR-ARFI allowed mapping the distribution of the radiation force at the focus of the array. After the recording of the MR phase signals for different US emissions, the proposed adaptive focusing technique was able to recover the spatial distribution of the phase aberrations. Total acquisition time for 384 ultrasonic emission channels was 2 hours. Conclusion: Those first results in clinical MR at 1.5 T show that adaptive focusing of a human transcranial brain HIFU system can be achieved within reasonable time under MR guidance for aberrator layers as strong as human skull. Ongoing work is aiming at accelerating the acquisition in order to reach acceptable durations for in vivo protocols.