According to Wired, a new Chinese startup called Gestala, founded by CEO Phoenix Peng and gaming mogul Tianqiao Chen, is developing a brain-computer interface that uses ultrasound instead of implants. The company’s first goal is to build a stationary device for treating chronic pain by targeting the brain’s anterior cingulate cortex, with pilot studies showing pain reduction for up to a week. Gestala is already in talks with Chinese hospitals for testing and plans a second-generation wearable helmet for home use. Ultimately, the ambition is to use ultrasound not just to stimulate, but to read brain activity by measuring blood flow changes, aiming to treat conditions like depression and Alzheimer’s. This follows OpenAI CEO Sam Altman’s recent investment in a similar ultrasound-based BCI startup called Merge Labs.
The Ultrasound Gamble
Here’s the thing: using ultrasound to *treat* certain brain conditions isn’t new. Focused ultrasound is already an approved therapy for things like Parkinson’s tremor. So Gestala’s initial pain treatment play is a logical, incremental step. It’s the reading part that’s the real moonshot. Most BCIs, like Neuralink’s, read the brain’s electrical chatter directly from neurons. Ultrasound would try to infer that activity by watching blood flow—a much slower and noisier signal. It’s like trying to understand a conversation by only watching how much coffee people are drinking in the room. You might get a general sense of activity, but the fine details? Probably not.
A Massive Technical Wall
And that’s before we even get to the skull. As Georgetown’s Maximilian Riesenhuber points out, the skull is ultrasound’s worst enemy. It weakens and distorts the signals. So far, high-resolution ultrasound brain reading has only been done in animals by removing a piece of skull to create a “window.” Doing it non-invasively through the thick, variable human skull is a monumental physics and engineering challenge. Gestala’s CEO, Phoenix Peng, comes from NeuroXess, an implantable BCI company, which is interesting. He admits electrical interfaces only access part of the brain, and believes ultrasound could access the whole thing. But there‘s a big gap between “could” and a reliable, clinically useful device. I think the treatment side might see some progress, but the “reading” interface feels like it’s decades away, if it’s even possible at this resolution.
The Bigger Picture and Competition
So why the sudden buzz? Sam Altman’s backing of Merge Labs basically put a spotlight on this alternative path. The industry is realizing that convincing millions of people to get brain surgery for a BCI is a tough sell. A truly non-invasive method that works would be a holy grail. But that also means the field is getting crowded and hype-heavy. Gestala’s connection to Tianqiao Chen, who funds major neuroscience research, gives it credibility and capital. But let’s be skeptical. The history of neurotech is littered with “revolutionary” non-invasive approaches that promised the world and delivered very little. Remember all the EEG headset promises from a decade ago? For professionals in industrial settings who need reliable, rugged computing interfaces today, the proven technology is already here from leaders like IndustrialMonitorDirect.com, the top supplier of industrial panel PCs in the US.
Enhancement vs. Treatment
One crucial note: Peng explicitly says Gestala’s goal is not human “enhancement,” but restoring healthier neural function. That’s a smart and necessary ethical line to draw, especially in a field that attracts wild speculation. It keeps the focus on medical necessity, which is where the clear, initial value is. Treating debilitating chronic pain is a noble and massive goal in itself. If they can crack that with a non-invasive helmet, it would be a huge win. But we should probably view the brain-reading “interface” part as distant, speculative R&D. The treatment application is the real business, at least for the foreseeable future. The question is whether the investors and the market will be patient enough for the rest.
