Every tool that reaches a deep brain structure has come with a trade. Implanted electrodes reach the subthalamic nucleus and other subcortical targets, but only after surgeons drill through the skull and leave hardware in the brain. Non-invasive stimulators spare the patient surgery, but their fields scatter within a few centimeters (an inch or two), so they mostly reach the cortex and leave deep circuits untouched. During sleep, when deep structures orchestrate shifts between sleep stages, that trade has been a wall: reach the deep brain, or leave the sleeper alone.
A new device built by a University of Texas at Austin engineering and sleep research team is designed to slip through that wall. Called NEUSLeeP, it is a soft, skin-attached patch that pairs a focused ultrasound array with EEG electrodes, letting it stimulate a deep brain target and listen to the sleeping brain at the same time. In a 28-participant trial reported in Nature Communications, wearing the patch overnight lengthened REM sleep by 4.6 percent and cut the time to the first REM period by 24 percent.
The patch
NEUSLeeP is built like a small, flexible bandage rather than a headset. Its core is a concentric-ring ultrasound array whose transducer phases can be tuned to steer and focus the beam. Focused ultrasound passes through the skull and converges on a spot a few millimeters (a fraction of an inch) wide, deep in the tissue, without heating or cutting anything along the way. Around the array sit conformal hydrogel electrodes that record brain waves from the scalp, plus compliant interconnects that flex with the skin. Everything is embedded in a soft, bioadhesive substrate engineered to stay attached through a full night of movement.
The design solves a problem that has kept ultrasound out of sleep research: mechanical stability. A rigid transducer pressed against a sleeping head drifts, and a drifting beam cannot hold a deep target. Because the patch conforms to the skin and moves with it, the array stays in place for hours, and the electrodes keep delivering usable sleep staging data while the ultrasound runs. The patch therefore monitors sleep stages in real time while delivering stimulation, so the effect of the beam can be read in the same recording.
The deep target: subthalamic nucleus
The target is the subthalamic nucleus, a small, lens-shaped structure buried in the basal ganglia. It sits among subcortical structures, including the pedunculopontine nucleus and the substantia nigra reticulata, that research ties to transitions between wakefulness and sleep stages. Decades of clinical observation support the connection: Parkinson’s patients with implanted subthalamic deep brain stimulators have shown changes in sleep architecture and sleep-related cognitive outcomes, evidence that this tiny nucleus is entangled with the machinery of sleep.
Until now, no non-invasive method could interrogate it directly. Transcranial electrical stimulation, transcranial magnetic stimulation, and closed-loop auditory stimulation all suffer from poor spatial resolution and off-target effects, so they are effectively limited to cortical targets such as sensorimotor or prefrontal areas. Focused ultrasound is different: it can be focused through the skull at depth, offering some of the spatial precision of an implant without the surgery. The patch’s ring array adds steerability, so the focus can be aimed at a specific nucleus rather than a broad region.
What 28 nights showed
The trial enrolled 28 adults, including healthy sleepers and people with self-reported insomnia symptoms identified by PSQI screening, none of whom carried a formal clinical diagnosis. Each participant slept with the patch in place under conditions comparing active focused ultrasound stimulation of the subthalamic nucleus with a sham control. The patch’s own EEG electrodes scored the sleep, and the comparison came out in REM: REM duration rose 4.6 percent, and REM latency, the time from sleep onset to the first REM episode, fell 24 percent.
The trial also looked beyond sleep stages. Heart rate variability was measured before and after sleep as an index of autonomic regulation, and participants performed the Hariri task, which probes response time and impulsivity across emotional stimuli, to test whether the stimulation shifted emotion-related behavior. These measures were analyzed with group (healthy versus insomnia) and condition (active versus sham) as predictors, with pre-sleep values as covariates. The trial carries the ClinicalTrials.gov registration number NCT07190287, and the protocol was approved by the University of Texas at Austin’s Institutional Review Board under the Declaration of Helsinki, with written informed consent from all participants.
Why it matters
Roughly 14.5 percent of adults report sleep difficulties, and poor sleep is intertwined with depression, anxiety, and PTSD. The standard options remain cognitive behavioral therapy and pharmacotherapy, both effective but both carrying residual effects that patients feel the next day: persistent drowsiness, fatigue, memory and mood impairment, and shortened attention.
NEUSLeeP points to a different path. As a research instrument, it is a causal tool: it can modulate a specific deep circuit during natural sleep and read the consequences in the same recording, the kind of experiment that has been out of reach without surgery. As a therapy, a wearable patch that nudges a sleep-regulating nucleus could eventually offer relief without the systemic effects of a drug. The REM findings are modest, but the platform, not the effect size, is the story: a non-invasive, spatiotemporally precise way to write to deep brain circuits overnight.
Limits
The results are early stage and should be read as a demonstration. One trial, 28 participants, and a mixed sample of healthy sleepers and self-reported insomniacs without clinical diagnoses: the numbers do not yet establish efficacy in insomnia disorder. The mechanism also remains unsettled. Low-intensity focused ultrasound can excite or suppress neural tissue, but how its mechanical effects translate into the changes seen here, and why subthalamic stimulation shifts REM in this direction, is not fully understood. Long-term effects of repeated nightly exposure, and the durability of any benefit, have not been tested. A patent application covering the work has been filed, and the device is years away from any clinical use.
Source
Tang KWK, Baird B, Moscoso-Barrera WD, et al. Skin-attached bioadhesive patch enabling ultrasound deep brain stimulation and real-time electrophysiological monitoring for REM sleep enhancement. Nature Communications. 2026;17:5570. doi: 10.1038/s41467-026-73787-6. ClinicalTrials.gov identifier: NCT07190287.