Economic Observer reporter Zhang Ling and Liu Xiaonuo

In early August 2026, Wang Zhiming, a paraplegic patient, walked step by step from his doorstep with his hands on braces, chatting with his twin daughters as he walked.

During this period, he occasionally posts videos of himself walking online. Many patients came to ask him how he stood up again.

Six years ago, Wang Zhiming's lower limbs were paralyzed due to a car accident, and he has been bedridden for a long time, with urinary and fecal incontinence. In May 2025, he participated in the clinical trial of the epidural brain computer interface "Beinao-1" jointly conducted by Xuanwu Hospital of Capital Medical University (hereinafter referred to as "Xuanwu Hospital") and Beijing Institute of Neuroscience and Brain like Research. After surgery, his spinal cord injury rating improved from the most severe AISA-A to C, and he was able to stand and walk with the assistance of a walker, restoring his ability to urinate and defecate independently.

The brain computer interface not only helps some paralyzed people stand up again, but also attempts to help blind people see the world again.

In May 2026, Chen Yiqi, an ophthalmologist at Zhejiang Provincial People's Hospital, performed retinal brain computer interface implantation surgery for Lingling. After nearly three months of rehabilitation training, Lingling, who is almost completely blind and only has a slight sense of light in strong light, can now recognize some Chinese characters and objects.

At the end of July, Lingling saw Chen Yiqi again and said to him, "It's quite happy to be able to see things now

Nowadays, a technology that once belonged to the science fiction world is entering the real human body on a large scale. Different technological approaches are attempting to bypass damaged neural pathways and re-establish the connection between humans and the outside world, treating diseases such as spinal cord injury, stroke, blindness, and coma.

However, the results of clinical trials are not always as desired, with some achieving significant improvements while others have limited effectiveness.

Wang Yihe, a neurosurgeon at Xuanwu Hospital, found that many patients believe that having a brain computer interface can restore them to normal. This is basically impossible. Besides, not all patients have an effect

He compared the current and future of brain computer interface technology to the "big brother" and smartphones in the field of communication. Patients can choose to receive a 'mobile phone' now or wait for the future.

Stand up again and see again

At the time of the car accident, Wang Zhiming was only 24 years old. A overturned semi-trailer pressed him under the car, causing him to lose consciousness below his belly button. He received traditional rehabilitation treatment such as acupuncture and moxibustion, electrotherapy and massage for two years in the local area, without obvious effect. The doctor told him that this is a complete spinal cord injury and he can only live in a wheelchair for the rest of his life.

But he is unwilling to give up. He asked his friends if there were more advanced medical technologies and also searched for information online. In 2025, he brushed a video from Huashan Hospital affiliated to Fudan University on Tiktok, and a paraplegic patient was able to grasp bricks and dumbbells by hand six months after implantation of brain computer interface. He thought, perhaps he could also stand up again.

He asked his brother to bring the film and diagnosis certificate to Xuanwu Hospital for consultation. After evaluation, he meets the clinical trial criteria. His hometown is 1500 kilometers away from Beijing. The doctor called him at 10 am, and an hour later, his wife drove him away.

To this day, Wang Zhiming still remembers clearly that when he woke up from the brain computer interface implantation surgery, his first words were to ask his wife, "Can you take a look, can my leg move

After more than ten days of surgery, Wang Zhiming began a six-month rehabilitation training. The sensation brought by electrical stimulation was like an electric current passing through the body, "as if nerves were connected," and that was the first time he felt his leg after being injured.

Wang Zhiming is the world's first patient to receive combined treatment using an epidural brain computer interface, spinal cord electrical stimulation, and exoskeleton robots. During training, he needs to generate action intentions through a brain computer interface. For example, when he imagines lifting his left leg in his brain and successfully transmits his thoughts, a current will flash through his left leg, and the exoskeleton will take a step forward with his left leg.

During training, he must be highly focused and clear his distractions, otherwise he may not be able to step out. At first, he thought about lifting his leg, but it didn't move. He had to think about it many times before taking a step. As the number of training sessions increased, he gradually became familiar with the control method. By the third month, he was able to take three steps at once.

Lingling, who is 40 years old this year, is a patient with retinitis pigmentosa and has been blind for nearly 20 years. Many people ask her, who is the first person she wants to see after the surgery? Her answer is son. She hasn't seen her son's appearance yet.

One of the reasons why the doctor chose her is that she is relatively young and has an optimistic personality. Brain computer interface is not about allowing patients to regain their original vision, but about giving up accumulated visual experience and rebuilding a set of visual cognition, which requires continuous learning and training.

Half a month after the surgery, Lingling started training and went to the hospital for 3 to 4 days a week, at least half a day each time. At first, she could see bright light, and gradually she could see formed and continuously existing shapes. Now, she can see some Chinese characters and distinguish some objects.

But there is still a long way to go before truly changing our lives. Chen Yiqi said that Lingling is still easily disturbed in complex scenes. According to the team plan, if she continues training for six months or even longer, she may further acquire the ability to walk, carry things, and pour water in complex environments.

To be honest, I don't have too much hope for her visual recovery. I just hope she can live independently, walk without bumping into people, and eat without feeding, "Chen Yiqi said." For many patients, brain computer interfaces may be the last resort

Restoring to a normal person is unlikely

Not everyone is as lucky as Wang Zhiming and Lingling.

On August 18th, a neurologist received a video from a rehabilitation hospital showing a hemiplegic patient holding onto a hallway handrail with their right hand and clenching their left hand into a fist, walking forward step by step, approximately every two seconds. More than a year ago, he performed brain computer interface surgery for this patient. Prior to surgery, the patient had grade III hemiplegia and required assistance while walking. They were unable to grasp spoons and were unable to take care of themselves.

After the surgery, dozens of paralyzed patients approached this doctor, hoping to undergo brain computer interface surgery, but he did not agree to them.

The doctor said that the medical team originally hoped to restore the patient's basic self-care ability through three months of training because the signal collection was not ideal enough for last year's patient. Nowadays, he can walk independently with the help of handrails, but the improvement in hand grip is limited, and he can only grasp thicker objects. The patient has shown significant improvement, but it is still slightly worse than we estimated. He cannot fully take care of himself

Even though they know the results may not be ideal, there are still people willing to try.

In April 2024, Lin Meng's mother suffered a sudden cerebral hemorrhage. After waking up, the mother's language ability was impaired and she could only speak short words, and her right limb was completely immobile. After rehabilitation treatment, the mother's cognitive function gradually recovered, but the right half of the body remained immobile.

In September 2025, Lin Meng learned on Xiaohongshu that brain computer interfaces may enable patients to control external devices through consciousness, and even restore their mobility and play games. She thought it was amazing, so she took her mother to a top tier tertiary hospital in central China to participate in a clinical trial. She hopes that her mother's right hand and leg can move again, and even walk again.

This experiment did not implant a chip into the mother's skull, only required wearing a brain computer interface device to collect brain signals, and trained by imagining movements such as lifting hands and kicking legs. The original plan was to train for 15 days, but on the 13th day, Lin Meng decided to withdraw because her mother's body had not shown any significant changes.

Lin Meng told the Economic Observer that if re elected, she is still willing to take her mother to participate in this experiment because non-invasive brain computer interfaces will not cause harm to the body at least.

In the view of the aforementioned neurologist, these ineffective or unexpected results expose two shortcomings of the brain computer interface: first, the collection of brain signals is still incomplete, and second, the conversion from brain signals to skeletal muscle movements is not yet mature enough.

At the end of 2025, Xuanwu Hospital will open a brain computer interface screening clinic. There are three to four hundred patients with spinal cord injury in the clinical trial database of Xuanwu Hospital. After basic data screening, initial screening through video conferencing, detailed outpatient evaluation, and discussing risks and benefits with patients, only about 10 people were ultimately enrolled.

Wang Yihe goes to the brain computer interface screening clinic every week. In addition to conducting preliminary clinical trials and popularizing science, he also needs to do one thing: reduce the expectations of patients and their families.

Special Surgery

Wang Yihe has been involved in brain computer interface work since 2023 and has participated in nearly 20 surgeries, including Wang Zhiming's surgery.

Wang Zhiming underwent a combination surgery. The doctor first implants electrodes outside the dura mater, then flips over to perform spinal surgery and places an electrical stimulator. The entire process lasts for seven to eight hours. After surgery, the equipment needs to be connected to the lower limb exoskeleton to form a closed-loop system of "brain spinal cord limb".

The difficulty of the surgery lies not only in implantation, but also in connecting the devices of the three companies. Due to differences in software and ports, the three companies repeatedly debugged, opened ports, and tested parameters before surgery.

In addition to the epidural brain computer interface, Xuanwu Hospital is also exploring interventional pathways. In 2022, Ma Yongjie, a neurosurgeon at Xuanwu Hospital, implanted an interventional stent electrode into a sheep and completed an interventional brain computer interface animal experiment.

At the beginning, the success rate of animal experiments was not high, and it was possible that one out of three would die. ”Ma Yongjie said that after dozens of experiments, the team gradually became familiar with animal anatomy, instruments, and surgical pathways. In the first half of 2026, electrodes in a sheep's body continuously and stably collected EEG signals for 8 months. Pathological examination after the experiment showed that the electrodes had grown together with the blood vessel wall, but the blood vessels remained unobstructed.

In September 2025, the team began human research and has completed interventional brain computer interface implantation in 4 patients with severe subarachnoid hemorrhage, all of whom are in a coma state. The research mainly verifies the safety and signal acquisition stability of the device in the human body. Four surgeries did not result in any adverse events related to the instruments. The electrodes were implanted in an average of about 7 minutes and can be removed in about 1 minute after collection.

Brain computer interface surgery does not only occur in the brain. For Chen Yiqi, he is facing another equally delicate and fragile organ - the eyes.

The surgery performed by Chen Yiqi for Lingling was his first brain computer interface human surgery. The surgery lasts from 1pm to 6pm, which is longer than regular retinal surgery.

Lingling needs to implant a chip with 320 electrodes in her eye, which is attached to the retina and fixed near the macula with a magnetic fixator. The camera on the glasses captures external images and transmits the signal to the chip, stimulating the retina to produce vision. The difficulty of surgery is not putting the chip in, but placing it in the accurate position. When separating the choroid and sclera, there should be no bleeding, and when moving the chip, the retina should not be scratched.

For this surgery, the team made a 3D model based on Lingling's eye condition before the operation, simulating the fixed position of the chip and the surgical incision. On the day of the surgery, the assistant quietly bought a can of Red Bull for Chen Yiqi.

Before performing surgery on Lingling, Chen Yiqi had completed hundreds of animal experiments over a period of 12 years. He first explored the implantation path with pig eyes, then observed signal transmission with rabbits. Last year, he implanted a chip in a monkey and removed it a few months later. No retinal detachment or eye atrophy was found.

At the end of July, when seeing Lingling again, Chen Yiqi conducted a detailed examination for her. Her cornea and lens are in good condition, the chip position is normal, and it fits well with the retina.

Many people are curious, can the implanted brain computer interface still be removed?

Ma Yongjie explained that the difficulty of retrieving devices varies depending on the technology path: SEEG (Stereoscopic Electroencephalogram) is usually used for a short period of time and can be retrieved after collection is completed; Epidural devices can be left in place for a long time and can also be removed if necessary; Long term retention of intravascular stent electrodes may result in fusion with the vascular wall, posing a risk of removal.

The Moon Landing Program Between Two Ears

Zhao Guoguang, the director of Xuanwu Hospital, regards brain computer interfaces as a "moon landing plan between two ears", which cannot only implant devices, but also ensure the safe "return" of patients.

According to the technical roadmap, brain computer interfaces can be roughly divided into non-invasive and invasive types. Non invasive methods mainly collect EEG signals through the scalp; Invasive methods require surgical implantation of electrodes into the body, including vascular intervention, epidural implantation, and direct implantation into the cerebral cortex.

The closer the electrode is to the neuron from the scalp to the inside of the brain, the clearer the signal can be obtained, but at the same time, the risk of invasion also increases. Zhao Guoguang made an analogy: non-invasive brain computer interfaces are like listening to music through a wall, epidural implantation is like sitting in the audience, and cortical implantation is like standing on a musical instrument to listen.

We hope to obtain high-quality signal acquisition with minimal trauma, "Zhao Guoguang said. Based on this concept, in 2023, the team collaborated with the Hong Bo team from Tsinghua University to explore the technology of epidural brain computer interface implantation.

They collaborated to complete the first epidural brain computer interface surgery, and the subject, Lao Yang, had been paralyzed for 15 years. After the surgery, Lao Yang's hand functions such as grasping, gripping, and pinching have significantly improved. Now he can hold a small wine cup and drink on his own, as well as pick up his grandson.

Although the technical routes are not entirely the same, the surgery we performed on Lao Yang was more than two months earlier than Musk's first human implantation at Neu Ralink company. ”Zhao Guoguang said.

In Zhao Guoguang's view, promoting brain computer interfaces into clinical practice is not because the technology is cool enough, but because there are many problems that traditional treatment methods cannot solve in clinical practice. Taking spinal cord injury as an example, there are about 3.74 million patients in China, many of whom were in their prime at the time of injury. After acute phase treatment and long-term rehabilitation, some patients will still enter a plateau period of functional recovery, and long-term bed rest and loss of self-care ability also bring heavy burden to their families.

Now we are challenging the impossible, "Zhao Guoguang said. Brain computer interfaces provide a possibility to bypass damaged neural pathways. After spinal cord injury, the patient's brain can continue to produce motor intentions, but these signals cannot be transmitted to the limbs through the damaged spinal cord. Brain computer interfaces can read brain signals and re-establish connections through methods such as spinal cord electrical stimulation, which is equivalent to building a bridge between damaged neural pathways.

This attempt does not mean that the patient can recover to a normal state. But for people who have been paralyzed for a long time, from being completely unable to move to being able to stand and walk with the help of assistive devices, it is already possible to change their lifestyle and also change the care status of the entire family.

Zhao Guoguang believes that brain computer interfaces will still undergo multiple rounds of product iteration, with more real-world data needed to answer the number of channels, implantation methods, and applicable diseases and populations. He also hopes to conduct controlled studies in the future to further validate the clinical benefits of brain computer interfaces.

Don't treat people like guinea pigs

Scientists can be crazy, entrepreneurs can be bold, but clinical doctors must be cautious, "said Ma Yongjie." We cannot treat people as guinea pigs

Ma Yongjie said that as the only group that can personally touch the brain, neurosurgeons will have a greater reverence for the brain. There were researchers from companies and engineering backgrounds pushing him to move forward faster, but he insisted on completing the necessary steps and verifying their safety and effectiveness before continuing.

He is wary of the risks brought by the rapid rise of brain computer interfaces. Some projects may compress the preclinical verification process due to chasing hotspots, and local government's industrial promotion may also put pressure on hospitals and enterprises. Strict doctors will carefully carry out every step before reaching the human body, but there are indeed cases of excessive celebration and competition for the top spot

The threshold for brain computer interfaces to enter clinical practice is ethical review. Many experts in hospital ethics committees are not familiar enough with such cutting-edge technologies, and clinical teams need to spend a lot of time explaining the technical principles, surgical risks, and potential benefits.

Chen Yiqi's project has undergone multiple rounds of ethical review. He said that the ethics committee is most concerned about safety, patient benefits, and patient protection.

Whether brain computer interface technology can truly enter clinical practice in the eyes still requires further observation. Chen Yiqi introduced that currently Lingling's visual perception can continue to exist, and how to further improve visual quality still requires longer training and follow-up in the future.

It took Ma Yongjie's team more than 3 years to complete animal experiments before applying for human research and submitting instrument type test results, animal experiment data, and patient inclusion and exclusion standards to the ethics committee. Due to the fact that the enrolled patients are in a coma state, the team also needs to explain the informed consent of their families and the patient's rights protection plan.

We must keep our original intention, not cause additional harm to patients, and carry out exploratory work on the premise of safety as much as possible, "said Ma Yongjie.

Zhao Guoguang said that compared with traditional medical technology, what makes brain computer interfaces more special is that they lack sufficient experience in human application, and some long-term risks can only be gradually recognized through continuous exploration.

A neurologist from a top tier tertiary hospital told Economic Observer that compared to mature implantable medical devices such as stents and pacemakers, brain computer interfaces are still in the early stages, and issues such as implantation sites, signal acquisition, and long-term safety need to be further validated through clinical trials.

The doctor is concerned that it is the patients who ultimately bear the technical uncertainty. He is particularly concerned about the long-term liability of implanted devices. Unlike drugs, implanted devices may remain in the patient's body for a long time or even for life. If a company goes bankrupt, who will bear the responsibility for subsequent maintenance and withdrawal? He mentioned that some medical device companies in the United States will cover such long-term risks through commercial insurance, while the relevant mechanisms in China are still incomplete.

For those who have already accepted brain computer interfaces, these unanswered questions are actually implanted devices that exist in their bodies.

Now, Wang Zhiming, who has returned to his hometown, is able to take care of himself. He can go to the bathroom by himself, and the residual urine during rehabilitation training gradually disappears. After the car accident, the severely atrophied and noodle like thighs have now bulged again.

Recently, parents returned to their hometown to harvest crops, and their children also returned to their hometown. Every day, after his wife goes to work, he stays at home alone, washing up, buying groceries, cooking, and taking a shower. He exercises with a brace every day and occasionally goes to the park with his family to exercise.

He has two implanted devices in his body, which he finds sci-fi, but in daily life, he hardly feels them.

Every few days, he will share his recovery process in Tiktok, how to train, what changes have taken place in leg strength, and how the urination and defecation functions are. More and more patients are asking him about his brain computer interface.

A few years ago, he was afraid of meeting people, and when he went out in a wheelchair, the gazes of passersby made him feel uncomfortable all over. Now, he goes out almost every day. His biggest wish is to one day get rid of braces, walk normally, and pick up his child from school.

(At the request of the interviewee, Lingling and Lin Meng are given pseudonyms in the article)

(Author Zhang Ling)