Home-administered unfocused transcranial ultrasound in subjective cognitive decline and dementia: an open-label,...

Memory and cognitive impairments remain disabling, with few effective treatments. Low-intensity transcranial ultrasound (TUS) is a non-invasive neuromodulation approach with a favorable safety profile and preliminary evidence of cognitive benefit, with one hypothesized but unproven mechanism involving microtubule resonance and reassembly.

This open-label pilot study evaluated the safety and preliminary efficacy of the US Pro2000 unfocused ultrasound device in adults with memory or cognitive difficulties, a sample spanning subjective cognitive decline among self-administering participants to diagnosed dementia among caregiver-rated recipients, and described self-reported symptom trajectories.

In a within-participant design, adults with self-reported cognitive impairment, or their caregivers, applied the device at the right temporal window for 5 min three times daily (target 15 min/day), in cycles of 3 weeks on and 1 week off. Outcomes (GAD-7, CESD-5, Inoue battery, and caregiver-rated MDS 3.0) and adverse events were collected online. Mixed-effects models with age and device-use covariates assessed change over time, with False Discovery Rate correction within aim.

Twenty-seven participants enrolled and 26 initiated treatment. Across 2,174 daily assessments, no serious or unanticipated adverse events occurred; only headache (15.4%) and transient jaw clenching (3.8%) were possibly related, and none caused discontinuation. Among self-administered participants, depression (CESD-5) declined significantly (p_FDR = 0.016) and five of eight monthly self-report symptoms improved, including memory, word-finding, and problem-solving (all p_FDR ≤ 0.040); anxiety trended downward, and the objective cognitive battery, near ceiling at baseline, did not change. Caregiver-rated outcomes were stable. Device use showed no dose response, and sensitivity analyses were consistent.

Home-based self-administered low-intensity TUS was well tolerated, with only mild transient events, although adverse-event ascertainment relied on unblinded self-report and relatedness judgments were made by unblinded investigators. Efficacy signals were confined to self-report, while objective cognition and caregiver ratings were unchanged. The uncontrolled design, small subgroups, and reliance on self-report preclude causal inference. These findings support the safety and feasibility of TUS and justify an adequately powered, double-blind, sham-controlled trial.