Source published October 20, 2026: peer-reviewed randomized sham-controlled crossover pilot in npj Acoustics, with an institutional research announcement from NYU Langone Health.
What the study tested Ten outpatients with schizophrenia or schizoaffective disorder and persistent psychosis completed one active and one sham session in random order, one week apart. The active session used MRI guidance to direct low-intensity focused ultrasound toward the right globus pallidus interna, a deep structure within a circuit linked to psychosis. Symptoms were measured before, during and through seven days after each session.
The early signal Delusion ratings improved more after active stimulation than after sham. The adjusted treatment estimate was -1.56 points, 95% confidence interval -2.54 to -0.58, p=0.002, with Cohen's d=0.68. Auditory-hallucination ratings and cognition did not differ significantly. No serious adverse event occurred. Seven participants after active ultrasound and all 10 after sham reported at least one increased side-effect item, including fatigue, headache, gastrointestinal discomfort or a brief increase in psychotic symptoms.
Why the boundary matters This pilot does not establish a treatment. Technical problems prevented the planned fMRI confirmation of target engagement. The delusion scale was designed for brief ecological sampling and has not been validated as a clinically significant treatment-response measure. Six participants guessed the condition correctly, so expectancy cannot be excluded. Follow-up lasted only seven days, dosing was not individualized for skull transmission, and individual responses varied.
What comes next A larger study would need clinically validated outcomes, confirmed targeting, stronger blinding, prespecified safety analysis, repeated-session testing and longer follow-up. The current result is most useful as a carefully controlled signal that a deep target can be studied noninvasively.
Funding and interests The study was supported by the Doris Duke Foundation Fund to Retain Clinical Scientists, Brain and Behavior Research Foundation Young Investigator Grant 31725 and the Max G. Schlapp Mental Hygiene Fund. The authors declared no competing interests and stated that funders had no role in the study.
Disclosure This is a research summary, not a treatment recommendation or evidence of established efficacy. Prepared with AI assistance; PubMed, the full open-access article and the NYU Langone announcement were checked October 27, 2026. The image is an AI-generated conceptual comparison. It is not anatomy, a medical device, measured data or a treatment outcome.