A patient in a coma cannot press a button or speak, yet somewhere under the bandage the brain is still listening. Doctors have long wanted a window into that silence — one that does not require drilling through bone. A team in the south of China thinks the window is made of sound.
The electrode problem
Almost every headline brain-computer interface (脑机接口) today depends on electrodes touching or entering the brain. That gives crisp signals, but it also means surgery, scarring, and a clock ticking on how long the implant stays reliable. For the millions of people who might benefit from non-invasive monitoring — stroke rehab, coma prognosis, mood and sleep — electrodes are overkill or unavailable.
Functional ultrasound, or fUS, offers a different route. It sends harmless sound waves through the skull and reads the echoes of blood flow that track neural activity. No wire into the tissue, no incision, just a probe on the head.
What Guangdong's team built
At the Guangdong Institute of Intelligence Science and Technology (广东省智能科学与技术研究院), researcher Zhihai Qiu (丘志海) has spent the last two years arguing that ultrasound brain-computer interface (超声脑机接口, or uBCI) is the next core direction, not a side experiment. In a 2025 forum recap from the Tianqiao and Chrissy Chen Institute, Qiu laid out three technical paths that make uBCI work:
- fUS imaging (fUSI) for high-sensitivity monitoring of neural activity across wide brain regions.
- High-precision ultrasound neuromodulation to stimulate specific circuits without electrodes.
- A closed-loop system that combines imaging and stimulation so the device can read, decide, and act.
The promise is resolution. Where scalp EEG blurs the whole brain into a smear, fUS can see activity at roughly 100-micron scale and reach 15 mm deep — fine enough to watch distributed circuits rather than guess from the surface.
Shrinking the machine
The catch has always been size and cost. Research-grade fUS rigs are room-filling and priced like medical imaging suites. Qiu's group responded with mini-fUS, a miniaturized platform described in a 2025 preprint from the team: a compact transmit-receive board and high-speed link that images a freely moving mouse's whole brain at up to 3.57 Hz, with about 100-micron resolution and 15 mm penetration, using a GPU-accelerated beamforming pipeline.
If a mouse can wear it, the reasoning goes, a human version can be portable. That is the jump from "lab instrument" to "bedside tool."
The coma bedside test
The most striking real-world signal came in late 2025. Qiu's team, with Fudan University's Huashan Hospital (复旦大学附属华山医院) and others, published in Research (a Science partner journal) a study using fUS to capture auditory responses in comatose patients — people who had undergone decompressive craniectomy after severe brain injury.
The finding: task-related brain activity picked up by ultrasound tracked with how those patients recovered nine months later. In other words, sound waves at the bedside could hint at whether a seemingly unresponsive brain is still in there, and how it might fare. The team calls it the first proof that bedside functional ultrasound can detect faint brain activity in severe coma — a sensitive, portable, low-cost prognostic tool.
Where ultrasound fits among non-invasive tools
fUS sits in a spot other non-invasive methods miss. fMRI sees the whole brain but needs a giant magnet and cannot sit at a bedside; fNIRS is portable but shallow and blurry. Ultrasound reaches deep, resolves fine, and fits in a probe — the profile a brain-computer interface (超声脑机接口) needs to leave the lab. It will not replace an electrode for crisp motor control, but for monitoring and modulation across a ward it is the more practical starting point.
Why this matters beyond the lab
Most brain-computer interface (脑机接口) news is about letting paralyzed people move cursors. uBCI points at a quieter, larger market: continuous, safe, non-invasive brain monitoring for the critically ill, the recovering, and eventually the well. A device that reads and writes brain state through the skull lowers the bar from "candidate for brain surgery" to "patient with a probe on the head."
Honest limitations
uBCI is earlier than electrode-based brain-computer interface (脑机接口) on the road to products. The mouse work is preclinical; the coma study involved only a small number of patients and shows correlation with recovery, not causation or therapy. The field openly admits hard unsolved problems — decoding efficiency, stimulation specificity, system integration, and real-time performance. fUS also reads blood-flow proxies for neural activity, not the raw spikes an electrode catches, so it will not replace invasive links for fine motor control. Think of it as a complementary modality, not a replacement.
What readers can do now
- If a relative is in neuro-intensive care, ask the care team whether functional ultrasound (fUS) or evoked-potential monitoring is available for prognosis — it is spreading from research centers like Huashan Hospital (华山医院).
- Researchers interested in non-invasive brain-computer interface (脑机接口) should watch Qiu's group and similar fUS labs; the miniaturization papers are the ones to read first.
- Everyone else: treat "ultrasound brain-computer interface (超声脑机接口)" as a promising monitoring tool in trials, not a consumer gadget you can buy yet.