NUPIAO Reporter |
NUPIAO Editor | Liu Fangyuan
On July 3rd, Gestala, China’s very first ultrasound brain-computer interface startup, made a splash by announcing a 420 million yuan (about $58M) angel+ round . That pushed the total funding for this six-month-old company to a staggering 570 million yuan.
That same day, Gestala’s Shanghai HQ, tucked inside the “Brain Intelligence Hub” in Minhang District, officially opened its doors. And here’s the kicker—just a week earlier, on June 26, the company’s global HQ and first ultrasound BCI superfactory in Chengdu had already roared to life. So now they’re running a dual-headquarters setup: Chengdu drives R&D and manufacturing, while Shanghai takes the lead on AI and algorithms.
Chatting with NUPIAO and other media, Gestala founder and CEO Peng Lei explained that Shanghai, as China’s BCI industry and research powerhouse, brings together top-notch clinical resources and serious AI computing muscle. That’s a non-negotiable backbone for training Gestala’s models and getting all that data aligned.
Gestala was born on January 1, 2026, and it’s walking a totally different path from traditional electrical BCIs. No skull-opening, no electrodes jammed into the brain. Instead, it uses ultrasound to beam through the skull and gently nudge neurons, while reading signals by tracking the tiny blood flow changes in microvessels that neural activity triggers.
Peng Lei puts it simply: ultrasound tech gives us a whole new lens for understanding and interacting with the brain—it’s really a deep, deep fusion of neuroscience and AI.
Globally, this ultrasound route is still super fresh. The US counterpart is Merge Labs, co-founded by OpenAI’s Sam Altman, which surfaced this January and bagged a $250 million angel round.
Zooming out to the macro picture, BCIs are hitting a hypergrowth phase. McKinsey crunched the numbers and estimates the global market for BCI medical applications could hit $40 billion by 2030 and soar past $145 billion by 2040. The serious medical segment, focusing on central nervous system disorders, alone could be around $15 billion by 2030.
Meanwhile, a 2025 report from the China Academy of Information and Communications Technology points out that there are already over 800 BCI-related companies worldwide, with resources heavily concentrated in China and the US. Among the many tech branches, ultrasound is emerging as a fresh, exciting entry point.
For Peng Lei, this is his sixth startup rodeo. He previously co-founded Naohu Technology, an invasive BCI player in China, and even stepped in as CEO. So why did he make the leap from electrical to acoustic for his second swing at the BCI game? Peng told NUPIAO it boils down to three fundamental breakthroughs that ultrasound brings to the table.
First off, the possibility of whole-brain read and write. Peng explained to NUPIAO that the old electrical approach is physically boxed in—you can’t just poke electrodes all over the cortex from surface to deep. Sure, it’s killer at precisely capturing and decoding local neuron signals. But the brain is a complex system where the whole is way more than the sum of its parts. With its ability to penetrate, ultrasound has a real shot at exploring how the brain’s overall coordination actually works.
Second, ultrasound BCI covers a much wider range of indications, which means it can reach a way bigger patient crowd. Gestala’s first target out of the gate? Chronic pain.
Peng also let NUPIAO know that they’ll be pushing forward in parallel with mental disorders (think anxiety, depression, autism, addiction), stroke rehab, and gnarly neural conditions like Alzheimer’s. These have huge patient bases and a clinical cry for non-surgical options.
And third, the ultrasound route is way more tightly woven with the AI industry—and that’s actually the core reason Peng started Gestala in the first place.
When explaining why ultrasound and AI are such a tight couple, Peng pointed out to NUPIAO that traditional electrical BCIs deal with local electrical signals. The algorithms are pretty simple—basic filter models can handle stuff like moving a cursor. But ultrasound BCIs have to juggle multimodal data all at once—blood flow, electrical signals, MRI, you name it. The data is massive and complex, and you absolutely need beefy AI compute to make sense of it all.
More importantly, there’s a roughly one-second biological lag between neurons firing and blood flow changes—a phenomenon called neurovascular coupling. Peng says one of the core missions of Gestala’s NeuroAI innovation center in Shanghai is to train foundational simulation models that align these complex data streams, aiming to squeeze that signal delay down to under 0.5 seconds and dramatically boost interaction efficiency.
Peng really hammers home that neuroscience and AI drive each other forward. BCIs aren’t just passively feeding off AI compute; the new discoveries about deep brain mechanisms will also offer critical biomimetic paradigms for the next generation of AI to break through bottlenecks like continuous learning and hierarchical memory.
On the commercialization timeline, Peng told NUPIAO that Gestala’s first-gen BCI product for chronic pain is slated to debut by the end of this year. At the same time, the company will drop its first batch of clinical trial data, research papers, and the latest AI model updates.
Why use chronic pain as the beachhead? Peng explained they’re not trying to replace your average over-the-counter painkillers. Instead, they’re laser-focused on patients with severe pain who are already at risk of addiction, or those who can’t tolerate long-term high-dose meds because of complications.
From what he shared, the first-phase clinical trial will mainly screen for subtypes of neuropathic pain like postherpetic neuralgia, fibromyalgia, trigeminal neuralgia, and cancer pain.
And here’s a reality check: the first-gen product coming later this year won’t be some cute portable home gadget. It’s a helmet-like, cart-based serious medical device meant for hospital use under a doctor’s supervision. Each non-invasive treatment session lasts about 30 to 45 minutes. Afterward, pain scale scores can drop by roughly 50%, with the relief lasting one to two weeks—kind of like a sustained-release drug effect.
Gestala’s short-term regulatory goal is crystal clear: after unveiling the first product this year-end, they’ll kick off massive clinical evaluations and, within about a year and a half, aim to snag the world’s first Class III medical device registration for a non-invasive ultrasound BCI by the first half of the year after next.
Now, about those consumer-grade BCIs that seem to hit the market faster? Peng stressed to NUPIAO that in his book, BCI commercialization has to follow a strict order: serious medical device first, then home medical device, and finally consumer electronics. You can’t flip the script.
He believes that only after you’ve proven safety and efficacy in real hospital clinical settings, gathered high-quality multimodal data with clear pathological labels, and polished your models on that, can you truly have the ability to generalize to the general population.
“Skip the first two medical steps and jump straight into consumer apps, and you’ll often struggle to build a solid data moat. The technical ceiling and room for imagination will be pretty limited,” Peng said.
Even though the serious medtech path is a long, slow grind, Peng told NUPIAO he’s willing to spend 10 to 15 years on this. Ample capital backing gives Gestala the patience to crack the deepest tech barriers.
To him, BCI isn’t just a business—it’s the ultimate key for humanity to explore its own central nervous system. Only by sticking to long-termism can we reach the vision of “Better Brains for a Better Humanity.”
