Optimizing Parameters of Low-Intensity Focused Ultrasound for Pallidal Modulation in Huntington's Disease

This is not a traditional three-arm parallel trial but a Stage I/II dose-finding study using an adaptive design. Participants are dynamically assigned to three pulse repetition frequency levels (10 Hz, 50 Hz, 100 Hz) based on predefined rules, rather than fixed randomization. The study adopts a two-stage utility-based Bayesian optimal interval (U-BOIN) design: Stage I: Dose decisions are based solely on dose-limiting toxicity (DLT) incidence. Participants are enrolled in cohorts of three. Based on observed DLTs, the next cohort's dose is determined (escalate/stay/de-escalate). Stage II: When any dose group reaches 6 participants or the highest dose is explored, safety and efficacy data are integrated to calculate a utility value. Subsequent cohorts are assigned to the dose group with the highest utility value. The study stops when any dose group reaches 12 participants or total enrollment reaches 24. Not all three dose levels may be utilized. Allocation is not fixed a priori.