Brain-computer interfaces are missing a "Lobster Pool"! Tsinghua University's Gao Xiaorong responds to a reporter from the 21st Century Business Herald: The large-scale brain electroencephalogram model foundation is most likely to be developed in China.

EveryDay Business News reporter | Zhang Rui EveryDay Business News editor | Wei Guan Hong

Since this year began, warm government policy signals for brain-computer interface technology have been coming in frequently. In the Government Work Report, “brain-computer interface” was included for the first time.

On March 13, the National Medical Products Administration approved the world’s first invasive brain-computer interface medical device—the “Implantable Brain-Computer Interface Hand Motion Function Compensation System” of Borecon Medical Technology (Shanghai) Co., Ltd.—for launch.

During the “2026 Zhongguancun Forum Annual Conference” held from March 25 to 29, during the “Brain-Computer Interface Innovation Development Forum,” a clear statement was made by Deputy Minister of Industry and Information Technology Ke Qixin: brain-computer interfaces are at a critical stage of moving from technical R&D to large-scale application. They need to further pool strength and jointly promote brain-computer interfaces to accelerate from the laboratory to real-world applications.

During the forum, reporters from the Daily Economic News interviewed multiple experts and people from the industry around issues such as the commercialization of brain-computer interfaces, technical routes, and future prospects.

Policy empowerment: commercialization accelerates, entering the “brain renovation” phase

With “brain-computer interface” being written into the Government Work Report for the first time, does this mean commercialization will speed up?

In response, multiple interviewees said, “That is certainly the case.”

Gao Xiaorong, a long-term professor at Tsinghua University and one of the principal founders of the discipline of neural engineering and brain-computer interfaces, began conducting brain-computer interface research in China as early as 1998. He told Daily Economic News reporters that this means “we are entering the ‘brain renovation’ phase— the whole world is turning into an era where people want to ‘renovate the brain.’”

When talking about the launch of the world’s first invasive brain-computer interface medical device, he believes it is “very meaningful.” “It has already been 50 years since the concept of brain-computer interfaces was proposed. Finally, products have been implemented on the ground.”

Jie Fu, CEO (Chief Executive Officer) of Shanghai Jinxing Biotechnology Co., Ltd., told Daily Economic News reporters that only market demand can truly drive industry development. Once the application endpoint on the medical side is opened up, it will become the “low-lying area” that attracts all kinds of technology transformation and implementation. Naturally, various resources, capital, and technology will gather toward this direction. “If all parties cannot see a clear path to monetization in the long run, all the R&D work done so far will lack an effective exit. “We can see that the country has been very proactive and pragmatic in pushing this.”

In May last year, Capital Medical University’s affiliated Beijing Tiantan Hospital opened a specialized out-patient clinic for brain-computer interfaces. Zhao Jizong, an academician of the Chinese Academy of Sciences and a professor at Capital Medical University’s affiliated Beijing Tiantan Hospital, told Daily Economic News reporters in an interview that, “At present, the out-patient clinic for brain-computer interfaces is very popular. Dr. Yang Yi, who goes for out-patient visits, often can’t get off work on time once it’s scheduled.”

Zhao Jizong said that opening the out-patient clinic has two purposes: one is to recruit patients for research; the other is to prepare for future promotion, which requires establishing a case database. At present, it mainly targets three groups: hemiplegia, paraplegia, and ALS (amyotrophic lateral sclerosis).

“Unlike a regular out-patient clinic, it requires evaluating many issues, including family circumstances, income, and the relationship between spouses, etc.” Zhao Jizong said. “In the past, we didn’t pay attention to these, thinking that once the patient came, that was enough. In fact, the issues are quite complex. They’re not just medical problems. Long-term paralysis often leads to family problems, social issues such as getting impoverished due to illness, and so on.”

He said that opening the out-patient clinic means that brain-computer interfaces have entered the view of ordinary patients, but whether it can be done is another matter. It is still in the clinical trial stage, supported by research funding.

Ecosystem to be built: currently lack****EEG foundation model

Currently, AI (artificial intelligence) development is surging. Zhao Jizong believes that in the course of developing brain-computer interfaces, AI technology is needed. Adding AI can help accelerate equipment updates and iteration as well as post-implant training—for example, whether AI can be used to produce templates with stronger applicability so that patients with different conditions can use them.

In Gao Xiaorong’s view, what the brain-computer interface field lacks most right now is foundation model building. An ecosystem similar to CUDA (a parallel computing platform and programming model developed by NVIDIA) has not been set up yet. “It’s like building a ‘lobster pond’— once we build the ‘lobster pond,’ everyone can ‘raise lobsters.’”

Gao Xiaorong said that what we need to do now is exactly this foundation-building work. But currently, no one wants to do this kind of ‘dirty and tiring work.’ It requires processing massive amounts of data. “We started doing brain-computer interface competitions in 2010 and have accumulated a large amount of data. Now we are working on foundation models and basic computing power, and we will also cooperate with relevant institutions to invest resources in building infrastructure. Like the development path of large models, we need someone to lay the groundwork first.”

Gao Xiaorong said that building this “lobster pond” involves a lot of work. “Simply put, we need data, algorithms, and computing power, and also application scenarios. Only after preparing all of these can we build this ‘lobster pond’— that is, the foundation of an EEG foundation model.” Gao Xiaorong said that he believes the foundation model for EEG will most likely be created in China, because our work is relatively ahead of schedule. “Just like language models have a foundation model, EEG also needs a foundation model.”

Controversy over the route: “inclusive products** must be non-invasive****”**

Brain-computer interfaces are roughly divided into two categories: one is invasive, which requires surgically implanting electrodes; the other is non-invasive, which collects signals through external devices such as head-worn setups.

In Zhao Jizong’s view, non-invasive approaches are the easiest to推广. Invasive signals are of better quality, but they have high technical requirements, making them more difficult. In addition, implanted devices may cause immune reactions, fiber encapsulation, signal attenuation, and other problems over the long term.

“Many companies in China are doing external-type caps, but most of what they do is to improve sleep, help students concentrate, and similar applications. Doing something truly centered on motion function rehabilitation might be better, but the drawback is that the signal quality isn’t as good as invasive approaches.” In Zhao Jizong’s view, “the simplest is the best.” Whether it is semi-invasive or fully invasive, it still requires opening the skull. No implantation can be guaranteed to have zero side effects, and this also depends on individual circumstances.

The global share of non-invasive versus invasive brain-computer interfaces is roughly 8:2. Is it because non-invasive is not as difficult?

In response, Fu Jie said that it is not because the difficulty is smaller. The core of brain-computer interfaces lies in real-time “reading” and “writing” of signals. Currently, most companies are dealing with multi-modal data acquisition (reading) and neural modulation (writing), and these two parts are often separated. As industry heat increases, these directions are collectively referred to as the brain-computer interface track. “Currently, about 80% of companies are still in the stage of signal acquisition or one-way writing. But to achieve truly non-invasive brain-computer products with closed-loop control and personalized adjustment capabilities, I think the industry still needs to go through a relatively long development process.”

Fu Jie candidly admitted that she is more optimistic about non-invasive approaches. Because chronic brain health problems are becoming a global “silent epidemic.” The real value of non-invasive brain-computer interfaces is not that the technology is cool. It is that it responds to a social reality: children who are stuck dealing with poor attention, delayed sleep, and anxiety; middle-aged people who are wrapped layer by layer by pressure, insomnia, and comorbidities; and elderly people who go from sleeping poorly to neurodegenerative diseases, with the social care burden at 1:2.5.

“From everyone’s lifetime perspective, chronic brain diseases are unavoidable. Most of these chronic brain problems are not suitable for being solved with invasive methods. From the standpoint of economic feasibility and risk, the risk-reward ratio isn’t a good match. So non-invasive brain-computer interface solutions targeting chronic brain diseases can definitely become that beam of light.” she said.

In Gao Xiaorong’s view, what is needed now is inclusive products, and it cannot be said that only billionaires can “renovate the brain,” while others can’t “renovate” it. The world’s first implanted product that is currently on the market is still not an inclusive product. “Inclusive products must be non-invasive—everyone can afford them. Invasive is more expensive and more complex than non-invasive.”

Promising outlook: during the “Fifteen Five” phase, it may be rolled out across the country, but it still faces multiple tests

When asked how far brain-computer interface technology will develop during the “Fifteen Five” period, Zhao Jizong said that by the “Fifteen Five” stage, it is hoped it can be rolled out across the country, but it will definitely be carried out in qualified hospitals, not just any organization.

Zhao Jizong emphasized that brain-computer interface technology is still in the trial stage and is not a substitute for traditional treatment methods. Instead, it provides another rehabilitation pathway. The promotion of the technology also needs to address many issues, such as professional training personnel and standard-setting.

He gave an example: after device implantation, it requires training by professional personnel, and there is currently a shortage of such talent. At present, it is mainly computer professionals helping with decoding and training patients, and the training time is long. Patients don’t go home after just three to five days of hospitalization; they first need to learn how to operate a computer, and understand what different signals represent.

“Currently, our plan is—after surgery, they stay in the hospital for one month. After discharge, they live near the hospital for another two months, and only then can they return home. Because it is in the research stage, the number of patients is limited. It’s basically follow-up-based: if there’s any problem, they come back at any time to get it resolved. If it’s going to be rolled out nationwide, who will do this work? So now we can only do it one by one. It’s not because there’s no equipment— the instruments are there, implantation is also relatively straightforward. The issue is that the post-implant training work can’t keep up.” he said.

Zhao Jizong explained to Daily Economic News reporters that after extracting signals, it is necessary to analyze which signal corresponds to which action. Many signals have no value, or they are not signals of the dominant hand. To extract valid signals, it is currently still done by computer professionals, who guide patients to move a cursor on the screen, tell them “higher, lower, left, right,” and then adjustments are needed. “The adjustment process is the training process— training them how to move.”

The reporter learned that because among patients coming to out-patient visits, groups such as farmers and workers account for a relatively high proportion, these patients’ rehabilitation training must first learn how to use computers.

Does the training necessarily have to be done through a computer? In the future, can it be replaced with a phone? Gao Xiaorong said that computers and phones should not be too different. In the future, it will definitely become phones. “Our lab is already developing toward glasses.”

In addition, the issue of funding is also very important. Zhao Jizong mentioned that last year, the United States said patients’ average cost per person was 5,000 USD. “I think that was very confidence-inspiring. That translates to about 30,000 to 40,000 RMB, and Chinese patients can still accept it.” But this year, the United States mentioned close to $50k. “At this price, it is still a high barrier for ordinary patients.”

For when large-scale commercial applications might appear, Fu Jie believes the key still needs to go back to the brain-health issue or the disease itself. Build a scientific dialogue logic between clinical and approval authorities, and prove what advantages this technology has over existing treatment methods— for example, a past therapy is effective for 50% of patients; now it can be improved to 75%, and in the future it may reach 90%. “I think the limiting step lies in approvals themselves. The market has already opened the door, but it doesn’t mean you can directly fly to the commercial end point. You still need to validate it step by step with solid clinical data.”

Cover image source: EveryDay Economic News media resources database