China's First Invasive Brain-Computer Interface Clinical Trial Marks a Milestone with 'Shanghai Solution' Outperforming Neuralink

The Life of the First Subject in China’s Invasive Brain-Computer Interface Clinical Trial Is Undergoing Changes—For Over Two Months, He Has Played Chess and Racing Games with His Family Using Only His Thoughts, with “Brain-Control” Efficiency Comparable to That of an Average Person Using a Computer Touchpad. Amid This Harmony, He Has Gained Deeper Insights into “the Happiness of a Family”: Being Together and Engaging in an Activity as One Is True Happiness.

For China’s brain-computer interface (BCI) field, this represents a veritable milestone. From non-invasive to semi-invasive and then to invasive approaches, Chinese scientists have maintained their relentless pace of exploration. China has become the second country globally, after the United States, to conduct clinical trials of invasive BCIs, with multiple performance indicators surpassing those of Neuralink, the BCI company founded by Elon Musk.

The Shanghai-based Center for Excellence in Brain Science and Intelligence Technology of the Chinese Academy of Sciences (hereinafter referred to as the Center for Excellence in Brain and Intelligence), in collaboration with Huashan Hospital Affiliated to Fudan University and relevant enterprises in Shanghai, has accomplished this groundbreaking feat. For Shanghai, which is building a global hub for scientific and technological innovation, this achievement provides a “Shanghai Solution” for the future industry of brain-computer interfaces.

["Neuralink Has Not Surpassed This Technology"]

Compared with Musk’s brain-computer interface company, Neuralink, what are the advantages of the “Shanghai Solution”?

“On the same technological track, our electrodes are ultra-thin and flexible, with dimensions comparable to a single cell—approximately one-hundredth the diameter of a human hair—and their softness is on the order of intercellular forces,” introduced Zhao Zhengtuo, a researcher at the Center for Excellence in Brain Science and Intelligence Technology. This neural electrode, featuring the smallest size and highest flexibility worldwide, is only 1/5 to 1/7 the size of those used by Neuralink, yet it is a hundred times more flexible. This allows brain cells to remain virtually “unaware” of the presence of a foreign object, thereby minimizing damage to brain tissue to the greatest extent possible. This innovative technology represents a global first in achieving long-term implantation in brain tissue without immune scarring, has gained widespread recognition from international peers, and has become an ideal tool for the long-term, stable measurement of neuronal activity.

In this regard, MIT Technology Review stated, “Neuralink’s technology is advanced, but it has not surpassed this technology.” Science commented that it “has the potential to build complex brain-computer interfaces.”

“We also have advantages in miniaturized system integration; the implant’s size is extremely compact, being only as large as a coin and half the size of Neuralink’s product,” introduced Li Xue, a researcher at the Center for Excellence in Brain Science and Intelligence Technology. The implant has a diameter of 26 millimeters and a thickness of less than 6 millimeters, making it the smallest brain-controlled implant worldwide.

“Unlike Neuralink, which relies on surgical robots for implantation, our implant does not need to penetrate the skull during surgery. Instead, we simply thin a coin-sized area of the skull above the motor cortex to create a recess for embedding the implant, and then drill a 5-millimeter puncture hole in this recess for electrode insertion,” said Professor Lu Junfeng, a neurosurgeon at Huashan Hospital and the lead surgeon for the first clinical trial of an invasive brain-computer interface. He noted that while the surgical procedure itself was a “first,” this minimally invasive approach is “routine” for surgeons and offers a “high degree of surgical friendliness” for patients.

All brain-computer interface systems seek a balance between surgical friendliness and signal quality. Signal quality determines the “ceiling” of application scenarios, while surgical friendliness represents market acceptance and scalability.

["Two-Way Commitment"]

Currently, there are three main technical approaches to brain-computer interfaces: non-invasive, where electrodes are placed outside the skull; semi-invasive, where electrodes are implanted beneath the skull but outside the cerebral cortex; and invasive, where electrodes are implanted into the cerebral cortex to directly contact neuronal cells.

“The human cerebral cortex is organized into six layers and contains 86 to 100 billion neurons, akin to a stadium with 86 to 100 billion spectators distributed across six tiers of seating,” Li Xue explained by analogy. “Non-invasive methods capture audience sounds from outside the stadium; semi-invasive approaches involve installing audio recording devices on the roof inside the venue; whereas invasive techniques entail inserting microphone-equipped probes into the spectator stands, thereby enabling clear auditory access to each individual spectator within that region.”

From non-invasive to semi-invasive and then to invasive, the signal quality of brain-computer interfaces gradually improves, but the technical difficulty also increases accordingly. Due to the ability to collect high-quality single-neuron signals, the invasive technical path is becoming a key focus in the technological competition among major powers.

Why Have the Courage to Choose the Most Challenging Technical Route? “Choosing this path was not a whim, nor merely an act of courage; it stemmed more from our interest in and recognition of the value of this direction,” Li Xue told reporters. Ten years ago, when they first encountered brain-computer interface (BCI) technology, they quietly resolved to focus on this area with immense application potential, tackling the most thorny issue—building the interface between the brain and machines—which is considered a critical bottleneck.

“The real turning point was my decision to come to Shanghai and join the Center for Excellence in Brain Science and Intelligence Technology,” Zhao Zhengtuo told reporters. After completing their studies, the two post-90s scientists embarked on a “mutual commitment” with the Center for Excellence in Brain Science and Intelligence Technology.

Mu-Ming Poo, an academician of the Chinese Academy of Sciences and Academic Director of the Center for Excellence in Brain Science and Intelligence Technology (CEBSIT), told reporters that as early as 2019, when Zhengtuo Zhao had just obtained his Ph.D. in Biomedical Engineering from the University of Texas at Austin, CEBSIT extended an invitation to him, recognizing the importance of brain-computer interface (BCI) technology and the shortage of relevant domestic talent. Zhao had long admired CEBSIT, a major hub for brain science research in China. In his early years, he had read an article by Poo published in Neuron regarding the China Brain Project, which deeply impressed him with its mention of large-scale neural activity recording tools. “It not only solidified my research direction but also planted a seed in my heart: working under the leadership of such a scientist would be a rare opportunity for young researchers.” Upon completing his postdoctoral research in the United States at the end of 2020, he eagerly joined CEBSIT, becoming the principal investigator for the project on “Brain-Computer Interface Technology and Its Clinical Applications.”

After earning her Ph.D. from the Department of Electrical and Computer Engineering at Rice University in the United States in 2020, Li Xue was selected for the inaugural cohort of the Young Investigator Program at the Center for Excellence in Brain Science and Intelligence Technology—appointed directly as a principal investigator without postdoctoral experience, and leading her own independent laboratory.

When asked why he joined the Center for Excellence in Brain Science and Intelligence Technology, Li Xue told reporters, “The logic of the brain-computer interface (BCI) industry follows an inverted pyramid model. In addition to addressing engineering-oriented challenges at the front end, such as neural interfaces and systems, the depth and breadth of BCI technology applications are determined by neuroscience’s understanding and cognition of brain functions. The Center for Excellence in Brain Science and Intelligence Technology has substantial expertise and advantages in this area, providing guidance and support for technological R&D. This is precisely where our team holds a competitive edge over its peers.”

["Fast Track"]

In Shanghai, brain-computer interface research by two post-“90s” scientists has accelerated onto the “fast track.”

Do you remember the world’s first somatic cell-cloned monkey? It was born at the internationally leading non-human primate research platform of the Center for Excellence in Brain Science and Intelligence Technology. Prior to China’s first invasive brain-computer interface (BCI) clinical trial, the safety and functionality of the device were validated in experimental macaques on this platform—not only did no infections or electrode failures occur over nearly 600 days, but the animals also achieved brain-controlled typing after training.

What further energized the research team was that, after a period of stable operation, the implants in the macaques were surgically removed safely and reimplanted at the same cranial burr hole site, thereby demonstrating the feasibility of surgical revision for device upgrades.

Micro- and Nano-Electronic Fabrication Platform, which is rare among institutions in China primarily engaged in basic research. With the support of Shanghai Municipality, the Center for Excellence in Brain Science and Intelligence Technology has established an advanced micro- and nano-electronic fabrication platform capable of precisely processing and fabricating brain-computer interface devices at scales ranging from micrometers to nanometers, thereby accelerating the research and development process.

Leveraging an electrophysiological tool platform built on ultra-flexible electrodes, the Center for Excellence in Brain Science and Intelligence Technology has become one of only three research institutions worldwide capable of conducting large-scale single-cell recordings in the human brain, the other two being Massachusetts General Hospital and the University of California, San Francisco.

Real-time online decoding is a critical component of brain-computer interface (BCI) technology. The team at the Center for Excellence in Brain Science and Intelligence Technology has developed the only BCI system in China to have obtained a registration-type inspection report, capable of capturing single-neuron signals at the millisecond level, thereby enabling patients with "brain-control" capabilities. Subjects need only wear a cap to establish a "Bluetooth" connection with their implanted device.

The brain-computer interface system is designed for a service life of five years. Upon future regulatory approval and market launch, it is expected to significantly improve the quality of life for patients with complete spinal cord injury, bilateral upper-limb amputation, and amyotrophic lateral sclerosis (ALS).

[Version 2.0]

Now, Musk’s invasive brain-computer interface clinical trials have reached the fifth patient. After China’s first case, is there a version 2.0?

“Next, we will attempt to enable the first subject to use a robotic arm, allowing him to grasp and hold a cup. Subsequently, we will tackle the challenge of controlling devices such as robot dogs and embodied AI robots, thereby expanding his life boundaries—a problem that Elon Musk’s team has yet to solve,” introduced Li Xue, adding that the plan for this year is to complete another 3–4 clinical trials.

Version 2.0 of the brain-computer interface system is scheduled for release by the end of this year or early next year. It will feature 256 data transmission channels, four times the number in current clinical trial products, while its volume will be one-third smaller than that of existing devices. In addition to aiding motor control in paralyzed patients, it holds promise for restoring language function in individuals with aphasia.

Brain-Computer Interface: Not Just a Mechanical Interface, It Reconnects the World.

Decoding the “language” of patients with aphasia can achieve an accuracy rate of up to 97.5%, and paralyzed patients can type 90 letters per minute using their thoughts... Currently, brain-computer interface (BCI) technology offers a novel solution for functional reconstruction of the body and the treatment of neurological disorders, and in the future, it will serve as a foundational technology for achieving human enhancement and human-machine integration.

Can Carbon-Based and Silicon-Based Life Forms Achieve Symbiosis? According to Zhao Zhengtuo, invasive brain-computer interfaces (BCIs) are poised to propel humanity into an era of brain-machine integration, empowering the human brain with artificial intelligence. In the more distant future, as the efficiency of human-computer interaction improves significantly, BCI products could even become consumer-grade applications targeted at healthy individuals.

In January this year, Shanghai released the “Action Plan for Cultivating the Future Brain-Computer Interface Industry,” with the vision of establishing a global innovation hub for brain-computer interface products by 2030.

Zhao Zhengtuo felt more motivated. “If I were to describe the most profound takeaway from the past few years, it would be: never underestimate the power of a clear vision. Some once considered our goals overly idealistic, but with the right team and environment, it is indeed possible to bring them to fruition step by step.”

Source: Jiefang Daily, reporter Huang Haihua, 2025-06-14

https://www.shobserver.com/staticsg/res/html/web/newsDetail.html?id=928344