The global field of brain-computer interface (BCI) technology is witnessing explosive breakthroughs. In 2024, Neuralink, a company founded by American entrepreneur Elon Musk, announced the success of its first human implantation surgery, enabling the control of a cursor with thoughts, which attracted worldwide attention. Just one year later, China’s “Beinao-1” completed a landmark first human clinical trial, achieving more complex speech decoding functions. This not only marks China’s official entry into the “First-in-Man” era of BCI but also shows the world the differences in technical routes and the competitive landscape between China and the United States in this cutting-edge technology field.
“Beinao-1” is a semi-invasive BCI system developed by the Beijing Institute of Brain Sciences and Brain-Inspired Intelligence (CIBR) in collaboration with Beijing Xinzhida Neural Technology Co., Ltd. It has successfully helped a 67-year-old amyotrophic lateral sclerosis (ALS) patient express specific sentences such as “I want to eat” using brain electrical signals. Unlike Neuralink, which focuses on invasive electrode implantation into the cerebral cortex, the Chinese team has chosen a more moderate technical path—placing flexible electrodes outside the dura mater to achieve high-quality signal collection and decoding. This achievement has been widely reported by international media such as CNN and has triggered a global discussion in the neuroscience community on “how to balance safety and accuracy.”
Decoding the “Electrical Wave Code” of the Brain
“Obtaining information through the dura mater and then decoding it into specific words is very interesting,” said Maximilian Riesenhuber, a neuroscience professor at Georgetown University in the United States, in an interview with CNN, expressing great concern about the technical path of “Beinao-1.”
The human brain has approximately 86 billion neurons, which transmit information through complex electrochemical signals, forming the source of all thoughts, emotions, and actions. The core of BCI lies in establishing a direct communication channel between the brain and the external world, with its technical foundation deeply rooted in understanding the essence of brain operations.
What exactly is BCI? In the words of Li Yuan, Director of Business Development at Beijing Xinzhida Neural Technology Co., Ltd., its core principle is “capturing brain electrical signals and decoding intentions.” The activities of more than 86 billion neurons in the brain are ultimately conducted in the form of electrical signals. Whether it is eating, sleeping, emotional memory, all can be collected, transmitted, and decoded through engineering and technical means.
Invasive and semi-invasive are the engineering and technical means to achieve the above functions. From the perspective of technical classification, BCI can be divided into non-invasive (scalp EEG), semi-invasive (cerebral cortex surface, such as “Beinao-1”), and invasive (intracerebral tissue, such as Neuralink in the United States and “Beinao-2”) according to the depth of signal collection.
As an achievement promoted by the Beijing Municipal Government through the Beijing Institute of Brain Sciences and Brain-Inspired Intelligence, “Beinao-1” not only verifies the feasibility of the technology but also allows the world to see China’s unique choice in this field. Different from the United States, which focuses on the invasive route within the dura mater, China has chosen the semi-invasive path outside the dura mater. This route reduces surgical risks while achieving high-precision signal collection. As Riesenhuber observed, “Chinese technology has reached the level of the United States, Britain, and other countries in terms of sophistication.”
Li Yuan said in an interview with Global Times reporters, “It is difficult to simply judge which technical route is better, so we are promoting both routes simultaneously. From the perspective of clinical acceptance and practicality, the promotion speed of ‘Beinao-1’ in clinical applications will definitely be faster. As our first-generation product, its surgical method does not require opening the dura mater, the surgical risk is extremely low, and it can already help paralyzed patients caused by spinal cord injury and stroke improve their quality of life.”
The Beijing Institute of Brain Sciences and Brain-Inspired Intelligence and Beijing Xinzhida Neural Technology Co., Ltd. also have the invasive “Beinao-2,” which aims to achieve the highest precision of signal collection. Li Yuan said that currently, the entire industry is exploring the functions that “Beinao-2” can achieve but “Beinao-1” cannot. From the perspective of the final signal precision and the expandability of the product, it may eventually be necessary to place the electrode near a single neuron.
How to Balance Safety and Accuracy?
The significance of “Beinao-1” goes far beyond being the “first case.” Its multiple technological breakthroughs have enabled China to gain a position in the global BCI competition.
The leapfrog breakthrough in the number of channels is the most intuitive highlight. Previously, the maximum number of channels in international semi-invasive technology was 64, while “Beinao-1” has directly doubled to 128 channels. This means that more abundant brain signals can be collected, and decoding accuracy is greatly improved.
“We can confirm that its design of 128 electrode sites has reached the global leading level—before this, the highest record of similar international technologies was 64 electrode sites. Imagine the technical system of BCI as a smartphone, which includes many components such as a screen, chip, and battery. The former also integrates a variety of key technologies: from the electrodes responsible for signal collection to the chips for processing signals, and then to the communication units for data transmission. Each link needs to be precisely coordinated,” Li Yuan explained to Global Times reporters.
It is particularly important that since the device needs to be implanted into the skull, the requirements for the biological safety of its materials and the temperature rise control during system operation are much higher than those of ordinary electronic devices. When a mobile phone overheats, it can be turned off to cool down, but abnormal temperature of BCI may directly endanger the brain. Therefore, the reliability, safety, and power consumption balance of the system have become technical difficulties—both high-throughput signal processing must be achieved, and power consumption must be minimized.
Electrodes have attracted much attention because they are in direct contact with brain tissue and need to meet both biological safety and signal collection stability: they must not damage brain tissue and must avoid being recognized as foreign bodies by the human immune system and attacked. The previous generation of rigid electrodes often caused rejection reactions, leading to the formation of scar tissue around the electrodes, eventually losing the ability to collect signals, and long-term effectiveness was difficult to guarantee.
At present, “Beinao-1” places the electrode outside the dura mater, significantly reducing surgical risks. The flexible thin-film electrode material it uses solves the problem that traditional rigid electrodes are prone to cause immune rejection, ensuring the biological safety of long-term implantation. At the same time, through in vitro wireless power supply technology, the safety risks of internal batteries are avoided, clearing the obstacles for clinical promotion. Meanwhile, “Beinao-1” is also seeking a better alternative to internal batteries. The wireless minimally invasive implantation form has laid the foundation for industrialization and promotion.
BCI is Not a “Sci-Fi Show”
Multi-dimensional verification of clinical applications makes the future industrialization direction of China’s BCI highly imaginative. At present, “Beinao-1” has completed 5 human implantations, covering various indications such as spinal cord injury, ALS, and post-stroke hemiplegia.
The first patient was a young man with high cervical paraplegia, who has been implanted with “Beinao-1” for more than 5 months. For this patient, the core goal of “Beinao-1” is to help him realize the replacement and rehabilitation of upper limb motor functions.
From the perspective of motor function replacement, when the patient’s own limbs cannot move, they can control the robotic arm or computer cursor through brain control technology, completing operations without hand movements. At present, the patient’s control of the cursor has been close to the state of free target control, realizing “pointing where to go.” This breakthrough means that even if the hands cannot move, patients can control the “mouse” through the brain to manipulate the electronic world.
“In terms of rehabilitation, in addition to training him to use the brain-controlled cursor, we also equipped rehabilitation devices—the most commonly used is to paste muscle stimulation patches on the corresponding muscle positions. When the patient imagines ‘holding hands,’ the BCI receives the signal and sends instructions to the muscle stimulation device, prompting the muscles responsible for hand grip to contract, realizing the grasping action,” Li Yuan said.
Li Yuan said that clinical observations have found that such movements driven by the patient’s own endogenous signals can effectively promote limb rehabilitation. Such patients have received conventional rehabilitation treatment after injury (such as due to car accidents, etc.), but from the perspective of big data, they usually enter a platform period after about half a year of training, and subsequent progress is extremely slow, or even difficult to achieve significant results. After receiving “Beinao-1” auxiliary training, the patient’s upper limb motor score increased significantly at one month and three months: previously unable to hold a water bottle, now not only can do it but also can pick up a small sleeve and put it on another pole, completing actions that the hands could not achieve in the past.
The second patient case is an ALS patient who has basically lost the ability to speak. This is also a case that has attracted international attention. The Beijing Institute of Brain Sciences and Brain-Inspired Intelligence tried speech decoding on the patient. Three weeks after the operation, he achieved high-precision decoding of a limited vocabulary containing 62 words, and could spell sentences such as “I want to eat,” “I want to drink water,” and “help me find a doctor.”
However, there is still much room for exploration in this field: there is no precedent for fully implanted BCI used for Chinese speech decoding before. English, as an alphabetic language, is different from Chinese ideographic language. How to find an efficient decoding strategy to achieve unrestricted word decoding still needs in-depth research. In addition, ALS itself will continue to progress, and the area where the patient’s brain signals are generated may atrophy, affecting signal quality. Therefore, the final effect of open-level decoding is currently uncertain.
Li Yuan emphasized, “BCI is not a ‘sci-fi show’ but to really solve the patient’s problems. For example, for spinal cord injury patients, we skip the damaged nerves and let the brain signals directly command external devices or muscles. This is the core value of the technology.”
Exploring Industrialization
When the clinical breakthrough of “Beinao-1”, a more critical question emerged: when can this technology move from the laboratory to patients? Li Yuan’s answer is: “After registering for clinical trials, it is expected to be launched in 2-3 years.”
Compliance and approval paths are the first barriers to industrialization. As a Class III active implantable medical device, “Beinao-1” needs to pass strict safety and effectiveness verification. At present, “Beinao-1” has completed very early clinical research, and the next step will be to start registered clinical trials, which will take about 2-3 years.
“It’s like an exam. First, finish the ‘motor function reconstruction’ exam paper, get the certificate, and then expand to more indications such as speech communication,” Li Yuan explained, which is in line with the international medical device approval logic—first focus on single-scenario verification, and then gradually expand the scope of application.
In January this year, the “Action Plan for Accelerating the Innovation and Development of Brain-Computer Interface in Beijing (2025-2030)” was officially issued. The Action Plan deploys 15 key tasks from five aspects: technological breakthroughs, platform building, cluster cultivation, scenario construction, and standard formulation, and forms 3 safeguard measures.
In terms of standard formulation, the Action Plan mentions accelerating the establishment of testing and evaluation methods for core indicators, encouraging the establishment of testing and inspection standard systems for various application scenarios, and shortening product research and development cycles and inspection cycles. Improve the industrial standard system, accelerate standard formulation in terms of technology, data, application, ethics, and safety, and strengthen standard application. At the same time, strengthen research on BCI safety prevention technology and ethical governance, and build a safety service system.
Policies provide further support for the industrialization of technology. CNN also mentioned in the report “China is catching up to the US in brain tech, rivaling firms like Elon Musk’s Neuralink” that although China’s BCI technology started in the 1990s, 20 years later than the United States, which started in the 1970s, China’s progress is “rapid,” and the Chinese government has also given strong support. In 2024, the Ministry of Science and Technology released the first ethical guidelines for research in this field. At the local level, municipal governments in cities such as Beijing and Shanghai have provided support for brain technology enterprises, covering all aspects from research, clinical trials to commercialization.
However, localized substitution remains an unavoidable challenge for China’s BCI technology. At present, the chip of “Beinao-1” still relies on foreign mature solutions, and key components such as batteries also need to break through industrial-grade reliability problems. “It’s not that we can’t make it, but we need to balance the number of channels, power consumption, size, and yield rate,” Li Yuan revealed that domestic teams are already testing alternative chips, and 40nm to 100nm processes can fully meet the needs, but a lot of testing and verification are still required.
In the longer term, the industrialization of BCI also needs to solve the problems of cost and popularity. At present, equipment in the clinical research stage needs to be provided to patients free of charge, and enterprises need to bear high R&D costs. After mass production in the future, how to reduce prices and make it affordable for more patients will be a common issue faced by the industry.
Related topics:
Korea’s Personality Test Obsession: A Mirror to Inner Struggles
IVF Breakthrough: 8 ‘Three-Parent Babies’ Born Mitochondria Disease-Free
Chinese Team’s New TCM Efficacy Evaluation Method Published Internationally
