MRgFUS procedure

The procedure was carried out in an MR-guided focused ultrasound device with a 1024-element, phased-array transducer of 220 kHz center frequency (ExAblate Neuro; InSightec Haifa). Intraprocedural MR imaging was acquired for interim evaluations of the patient. Also, the real-time acoustic signal was monitored for adequacy of the spectral dose. Prior to the procedure, a stereotactic frame was affixed to the shaved head of the patient under local anesthesia. Patients entered the MRI, where the frame was coupled to the helmet transducer. Patients were mildly sedated and monitored during the procedure by an intensive care unit doctor.

A 3-Tesla MRI (Discovery 750 w, GE Healthcare, Milwaukee, Wisconsin) was used. MR images T1 weighted without contrast baseline, T2-weighted fast spin echo, and T2*-weighted gradient echo (GRE) were acquired for surgical planning and target selection (right parieto-occipito-temporal cortex) for BBB opening. Areas containing vessels and sulci within two contiguous MRI slices in each plane were spared to minimize the risk of bleeding. Patients received small intravenous boluses of weight-based microbubble contrast (Luminity^®, 4 μl per kg) immediately followed by the application of low-frequency FUS into the target. MR thermometry allowed real-time monitoring of tissue temperature in the sonicated region. Details of the procedure¹⁶ are summarized below. At each new target, power ramp sonications were performed during the microbubble injection in order to detect the lowest threshold for acoustic activity indication of putative cavitation²⁴. Subsequently, several sonications were performed at half of the detected power threshold. Sub-harmonic acoustic dose was monitored at each sonication, and the power was increased when less than optimal dose levels were achieved. For stage 1, sonication volumes were delineated by a rectangular spot of ~6 × 6 mm comprised of a 2-by-2 grid of spots with 3 mm spacing. Detailed information about sonication parameters is reported in Supplementary Table ¹. For stage 2, performed 2–3 weeks after stage 1, the same protocol was repeated at the original location as well as in an adjacent area. After each sonication, single-slice T2*-weighted images centered at the target location were obtained to monitor for microbleeding and hypointensities. Once the sonication procedure was completed, BBB opening was verified via gadolinium-enhanced T1-weighted images. Patients were then taken off the ExAblate bed, and transferred to the MR bed in order to perform brain MRI using the dedicated head coil. The after-treatment MRI protocol included whole brain T2*-weighted GRE, Susceptibility Weight Angiography (SWAN), and T1-weighted imaging after gadolinium injection. Following the clinical protocol used with FUS (thalamotomy, pallidotomy, etc.) in our center, patients were transferred to the hospital’s intermediate care unit to be monitored for a few hours.