Patient-specific effects on sonication heating efficiency during magnetic resonance-guided focused ultrasound...

During magnetic resonance-guided focused ultrasound (MRgFUS) thalamotomy for refractory tremor, high temperatures must be achieved and sustained for tissue necrosis. We assessed the impact of both patient-specific as well as procedure-related factors on the efficiency of acoustic energy transfer, or heating efficiency (HE).

Retrospective analysis of 92 consecutive patients (857 sonications) with essential tremor or tremor-dominant Parkinson's disease treated at a single institution. Temperature elevations at the target were measured for each sonication with MR thermometry. HE of each sonication was defined as the ratio of peak temperature elevation and the delivered energy. HE was analyzed with respect to patient skull features (area, thickness, skull density ratio [SDR]), computed from CT scans, as well as demographic and clinical variables (age, sex, diagnosis, and duration of symptoms).

Across the full range of sonication energies that can be delivered with current devices (up to 36 kJ), average sonication HE was diminished in patients with lower SDR. In individual subjects, there was a progressive loss in HE as sonication energy was titrated up throughout the course of treatment, with a more rapid decline in patients with higher SDR. This energy-dependent loss in HE was not related to procedural factors, namely, the number of previous sonications, or the cumulative energy deposited during previous sonications. In contrast to SDR, neither skull area nor thickness was an independent predictor of average HE or the rate of its decline with increasing energies. In 11% of patients, all of whom with SDR < 0.45, sonication HE fell below the threshold to reach 54°C even with delivery of maximum energy. In contrast, temperatures $$ 50°C could be obtained in all but one patient.

SDR is predictive of sonication HE, and determines patient-specific limits on the magnitude of temperature elevation that can be achieved with current devices. These data inform strategies for predictable lesioning in MRgFUS thalamotomy.