A Spiral Electrode That Moves With the Brain
Chinese researchers have developed a stretchable brain electrode designed to address a persistent problem in thought-controlled technology: implants can pull away from the tissue they are meant to monitor as the brain moves inside the skull. Their spiral structure can extend and compress like a spring, allowing it to follow those movements while putting less force on delicate tissue.
Contents
- A Spiral Electrode That Moves With the Brain
- Why Bending Alone Is Not Enough
- How the Spiral Reduces Force
- China's First Human Trial and Its Timeline
- From Computer Games to Paid Work
- Companies Prepare for Wider Use
- Policy and Medical Billing Take Shape
- Different Approaches in the Global Competition
- Durability, Replacement and Neural Data
- At a Glance
The work, led by Fang Ying at the Chinese Institute for Brain Research Beijing, appeared in Nature Electronics on February 6, 2026. The researchers tested a high-density array with 1,024 channels in macaque monkeys, demonstrating the collection of large numbers of neural signals and stability over an extended period.
The central advance is the difference between bending and stretching. A conventional flexible electrode can bend, but it cannot necessarily lengthen when its attachment points move apart. The new structure turns that stretching force into bending and coiling, reducing the tension that can pull an electrode out of position.
Fang's team encountered the problem during experiments with non-human primates approximately four years ago. She explained in an interview that the researchers observed a genuine risk of flexible electrodes retracting because of brain movement. That finding led them to investigate structures that could stretch without applying damaging force to surrounding tissue.
The development comes as Chinese research institutions and companies expand invasive brain-computer interface trials. These systems, commonly called BCIs, collect electrical activity from the brain and translate signals associated with intended movement into commands for computers or other machines. Their usefulness depends on electrodes maintaining a reliable connection with nearby neurons.
Why Bending Alone Is Not Enough
The brain moves roughly 100 micrometers with each heartbeat and also shifts with breathing and head motion. A micrometer is one-thousandth of a millimeter, so these movements are small, but they happen repeatedly around an implant that must continue recording precise signals.
An electrode anchored at one end to the skull and at the other to brain tissue experiences tension whenever those two points move relative to each other. A straight electrode may bend easily while still resisting a pull along its length. That resistance can make it withdraw from the tissue.
Electrode retraction drew international attention in 2024, when Neuralink reported that approximately 85 percent of the 1,024 flexible electrodes in its first human patient's implant had retracted within weeks of implantation. Losing contact with the intended recording sites reduces the signals a system can collect.
Repeated mechanical stress can also damage tissue and trigger inflammation. The brain may form scar tissue around the implant, a response known as glial scarring. That tissue separates the electrode from nearby neurons, progressively weakening the recorded signals until a device can become effectively unusable.
For invasive BCIs, this mismatch between a moving organ and an implanted device has been a stubborn obstacle to lasting performance. Better signal processing can help a system interpret the activity it receives, but a stable physical connection remains essential to collecting that activity in the first place.
How the Spiral Reduces Force
Fang's team redesigned the electrode's geometry instead of relying on straight or simply curved wires. The new electrodes use ultrathin spiral structures that accommodate movement through low-energy buckling, meaning the structure changes shape by bending and coiling as it is pulled.
Fang said stretching a conventional linear electrode by 100 micrometers requires approximately 4 millinewtons of force. The stretchable design requires only 37 micronewtons for the same movement, roughly one-hundredth as much force. The comparison puts the benefit in practical terms: the brain can move the new structure with far less resistance.
The spiral can both extend and compress, helping it follow the brain's rhythmic motion without transmitting damaging shear forces. Shear is the sideways force produced when adjoining surfaces move against each other. Around an electrode, that force can strain sensitive tissue.
The lower mechanical stress translates into less tissue damage and fewer immune responses at the implantation site. The electrodes are made from ultrathin flexible films that combine the ability to collect many neural signals with the ability to accommodate movement.
The 1,024-channel array tested in macaques matches the core channel scale of commercial systems while providing better mechanical integration with living tissue. A channel is a route for recording neural activity. More channels allow a system to collect signals from more recording points, but those points must remain in useful positions for their number to matter.
China's First Human Trial and Its Timeline
In March 2025, the Center for Excellence in Brain Science and Intelligence Technology at the Chinese Academy of Sciences, working with Huashan Hospital affiliated with Fudan University, launched China's first human clinical trial of an invasive BCI system. The trial made China the second country, after the United States, to advance invasive BCIs to the clinical stage.
The first participant was a man who had lost all four limbs in an accident involving high-voltage electricity 13 years earlier. He received the implant on March 25, 2025. The surgical team used high-precision navigation to position electrodes in the motor cortex with millimeter accuracy while minimizing trauma during the procedure.
The motor cortex is the part of the brain involved in directing movement. Recording activity there allows a BCI to interpret a person's intention to move and turn it into a command, even when an injury prevents the body from carrying out that movement.
The main developments unfolded across several years:
- Approximately four years before the February 2026 publication, Fang's team observed electrode retraction risks during experiments with non-human primates.
- In 2024, Neuralink reported substantial electrode retraction in its first human participant.
- In March 2025, China launched its first human invasive BCI trial, with the first implant placed on March 25.
- In June 2025, a second participant, Mr. Zhang, received the WRS01 wireless implant.
- On February 6, 2026, the stretchable electrode research appeared in Nature Electronics.
Within two to three weeks of training after his operation, the first participant could use his thoughts to control electronic devices, including racing games and chess programs.
From Computer Games to Paid Work
Mr. Zhang, the second participant, suffered a high-level spinal cord injury in 2022. After receiving the WRS01 implant in June 2025, he learned to control a smart wheelchair, command robotic dogs and carry out paid remote work verifying products for vending machines.
His work involved serving as an intern sorter and moving a computer cursor through neural signals. He became the first known BCI trial participant to achieve paid employment using only those signals, extending the technology's use beyond demonstrations and into a task with an income.
Mr. Zhang, the trial participant returning to work after his spinal injury, described what the opportunity meant to him:
It has been more than three years since the incident, and now I can finally work again.
The system processes neural signals, interprets movement intentions and generates control commands within tens of milliseconds, faster than the blink of an eye. WRS01 also uses optimized wireless communication protocols to achieve an end-to-end delay below 100 milliseconds.
Those measures describe related stages of the same interaction. Signal interpretation produces a command, while the end-to-end delay covers the wider process of delivering control to a device. A short delay helps cursor movement or wheelchair commands feel immediate and natural to the user.
The participants' activities show why stable recording matters. Games, mobility controls and remote work all require repeated commands. An implant that maintains contact with the intended neurons gives its processing system a more dependable stream of information to interpret.
Companies Prepare for Wider Use
Zhiran Medical, a company co-founded by Fang Ying, recently secured more than 300 million yuan in Series A financing to advance invasive BCI technology toward clinical use. Series A is a stage of company fundraising used to support further development and growth.
The company has developed fully implantable wireless systems supporting up to 1,024 channels. Their processors are about the size of a coin and reduce power consumption by 75 percent compared with earlier designs. Zhiran Medical has also established facilities in Beijing capable of manufacturing the stretchable electrodes at scale for clinical use.
StairMed Technology is pursuing another approach with ultra-flexible electrodes measuring one-hundredth the diameter of a human hair. Their small size is intended to make them barely perceptible to brain cells. The company plans trials at multiple centers involving 30 to 40 participants and is targeting market approval by 2028.
Shanghai Brain Tiger Technology has demonstrated flexible electrode arrays with 256 channels and information transmission rates comparable to Neuralink's published data. Transmission rate describes how much usable information a system can communicate, making it another measure of performance alongside channel count.
The companies are addressing different parts of the same practical challenge: recording useful signals while limiting disruption to brain tissue, then turning those signals into dependable wireless control. Financing and manufacturing capacity are also necessary to move devices from research production toward broader clinical testing.
Policy and Medical Billing Take Shape
China has included BCI development in recommendations for its upcoming 15th Five-Year Plan. That places the technology within the country's planning priorities as research groups and commercial developers work toward medical applications.
In March 2025, the National Healthcare Security Administration established specific pricing items for invasive BCI implantation and removal. These items create a framework for clinical billing and insurance coverage by identifying the procedures that hospitals may need to charge for.
Implantation and removal both matter because an invasive device involves surgery at the beginning of its use and may require another procedure later. A billing framework gives those activities a defined place in the medical payment system as clinical development continues.
StairMed's 2028 target provides a stated commercial milestone, while its planned trials involving 30 to 40 participants provide a next step toward that goal. The stretchable electrode research itself was tested in macaques; no date was given for a human trial of that particular design.
Different Approaches in the Global Competition
China's progress places its developers in direct competition with American companies, particularly Neuralink, over the commercialization of thought-controlled devices. The United States retains advantages in chip integration and complete system products, while Chinese teams have developed strengths in flexible electrode materials and techniques intended to reduce surgical trauma.
The Chinese electrodes described in the comparison have cross-sectional areas between one-fifth and one-seventh those of Neuralink's devices, with flexibility more than 100 times greater than competing products. Cross-sectional area is the size of the electrode's profile when viewed across its width, a measure relevant to how much space it occupies in tissue.
Neuralink's fully invasive system requires removing a portion of the skull and implanting rigid or semi-rigid electrode arrays, carrying higher surgical risks. Chinese teams have focused on small skull incisions and ultra-flexible probes as part of their approach to minimizing injury during implantation.
Lu Junfeng, a neurosurgeon at Huashan Hospital, explained the different BCI approaches through a sports analogy. A non-invasive device resembles a microphone outside a stadium, where it mainly hears crowd noise. A semi-invasive device resembles a microphone hanging above the field, where the sound is clearer. An invasive device resembles a microphone attached directly to the players, where individual speech can be heard distinctly.
The analogy explains why electrode placement matters. Moving closer to the source of neural activity improves access to detailed signals, while placing a device inside brain tissue creates demands for careful surgery and compatibility with that tissue.
WRS01 also benefits from China's strengths in telecommunications infrastructure. Its optimized wireless protocols help keep the full control delay below 100 milliseconds, connecting the work on electrodes and signal processing with the practical task of delivering commands quickly.
Durability, Replacement and Neural Data
The stretchable design addresses a mechanical barrier, but lasting compatibility with the brain remains a central concern. Acute tests show minimal immune response. Researchers still need continued observation of how the brain reacts to implanted foreign objects over years or decades.
Glial scarring makes that follow-up especially relevant. Even a small change around a recording site can affect the distance between an electrode and the neurons it monitors. Reducing force at the interface is therefore part of sustaining signal quality over time.
Device upgrades present another challenge. An implanted system cannot be replaced as easily as an external computer or phone. Replacing it requires additional surgery, making the durability of its electronics and its physical connection with tissue important to patients.
Neural data also require privacy and security protections. Invasive BCIs create direct pathways into the brain, bringing concerns about unauthorized access and misuse of the information they collect. Medical ethicists have called for strict standards for identity authentication, information encryption and protection of brain data.
Identity authentication checks who is allowed to use a system, while encryption protects information by encoding it. Both become important when a device interprets neural signals and sends commands to a computer, wheelchair or another connected machine.
Patients also need careful psychological screening as they adapt to machines that interpret their thoughts. The February 2026 electrode findings, the Chinese clinical trials and the companies' development plans together show progress across materials, surgery and practical control, while long-term observation remains necessary before invasive BCIs become routine treatments.
At a Glance
- Fang Ying's team published its stretchable brain electrode design in Nature Electronics on February 6, 2026.
- The spiral requires 37 micronewtons to stretch 100 micrometers, compared with approximately 4 millinewtons for a conventional linear electrode.
- A 1,024-channel array was tested in macaque monkeys, demonstrating large-scale recording and long-term stability.
- China's first human invasive BCI trial began in March 2025, with the first participant implanted on March 25.
- Participants have controlled games, a smart wheelchair and robotic dogs; Mr. Zhang also returned to paid remote work.
- Zhiran Medical raised more than 300 million yuan, and StairMed plans trials involving 30 to 40 participants with market approval targeted for 2028.
- China has introduced pricing items for BCI implantation and removal and included the technology in recommendations for its 15th Five-Year Plan.
- Lasting tissue compatibility, surgical replacement, neural data security and patient adaptation remain challenges.






