The short version
- Low-intensity focused ultrasound can change activity in deep parts of the brain without surgery, but a new critical review says it has not yet been shown to treat any condition.
- Most human studies so far have fewer than 20 people, many have no proper sham, and only 37% of completed registered trials have led to a published paper.
- The sound the devices make can itself change brain readings, so the authors want every trial to include a sham that sounds the same.
Low-intensity focused ultrasound can reach deep inside the brain and change what neurons do there, but no one has yet shown it treats any illness. That is the verdict of a critical review published in Frontiers in Neuroscience on November 10.
The technique is not the high-intensity focused ultrasound that burns a tiny spot in the brain to stop tremor, which is approved and which we covered for Parkinson’s last week. The low-intensity version uses far weaker sound, warms tissue by less than half a degree, and aims to nudge brain circuits for a while instead of destroying tissue. It has drawn attention because it can reach targets that otherwise need a surgically implanted deep brain stimulator.
What the review is
This is not a new trial. The authors, a team of neurosurgeons at Aviation General Hospital in Beijing with colleagues at Tianjin Huanhu Hospital and Beijing Tiantan Hospital, call it “a narrative critical review rather than a systematic review.” They searched six databases and ClinicalTrials.gov up to June 30, 2026, and read the laboratory, animal and human work together. A narrative review is a reasoned stock-take, not a pooled statistical answer, so its judgments are the authors’ own.
What it found
The physics works. In people, a burst pattern of ultrasound to the motor cortex leaves the brain more excitable for 30 to 60 minutes, and imaging and recordings show deep targets responding. The authors stress that this shows the sound reached its target. It does not show that anyone got better.
The clinical studies are small. A 2025 depression study treated five patients over five days and reported an average 60.9% drop in symptom scores, but with no sham group, so “placebo effects and regression to the mean cannot be excluded.” A double-blind Alzheimer’s pilot in nine patients found a benefit on one memory test at 24 weeks, while the comparison on its main endpoint was not significant. Parkinson’s studies of 10 and 13 people disagreed on whether movement scores improved. The tremor reports are case series of three to ten people. A 10-person schizophrenia pilot we wrote about on November 2 fits the same pattern: sham-controlled, which is a strength, and tiny.
There is a lot of activity behind these few papers. A registry analysis the authors cite found 177 low-intensity ultrasound trials worldwide by September 2025, a third of them in psychiatry. Only 37% of the completed ones had been linked to a peer-reviewed paper, which the authors read as a possible sign that negative results go unpublished.
The sound in the room
One problem is the sound itself. Pulsed ultrasound can be heard by the person receiving it, and that hearing can change brain readings on its own. A 2024 study by Kop and colleagues found that much of one inhibitory effect could be explained by that hearing pathway alone. Animal work in which hearing was removed still showed direct effects, so the authors do not think the whole field is an artifact. But many earlier studies used a sham in which the device is simply flipped away, which sounds different. The review says “active auditory controls must become a minimum standard.”
Safety
The authors report that the method looks safe at the doses used. An earlier review of 35 studies found mild side effects in 3.4% of participants (14 of 425) and no serious ones. Today’s limits, set by an international consensus group, cap the mechanical intensity and allow a temperature rise of no more than 2 °C. The authors note these are expert agreements, not hard safety boundaries.
Why it matters
The review’s own summary is the honest line: “LIFU is a promising but still largely investigational technology.” Sound that can reach the deep brain without surgery would be a real change for depression, pain and movement disorders. Getting there needs randomized trials big enough to detect a modest effect, with shams that sound like the real thing and follow-up of months, not days. The authors say no indication should be treated as established until those trials report. It is also worth watching which of the 177 registered trials publish, and which stay quiet.
