Blind Patients Perceive Shapes and Colors, Paralyzed Elderly Control Wheelchairs by Thought: Shanghai Brain-Computer Interface Conference Unveils Technological Breakthroughs

(Source: Shanghai Observer)

Today, Shanghai Xinhongqiao Research and Innovation Center has become a "showcase" for brain-computer interface companies and research teams from all over China. Some companies held new product launch events, multiple research teams participated in brain-computer interface competitions, and many enterprises appeared in the exhibition area, showcasing the interim achievements of Chinese researchers dedicating themselves to this future industry.

December 4th to 5th, the 2025 Brain-Computer Interface Conference themed "Brain Connects World·Intelligence Converges Shanghai" was held in the "Brain Wisdom World" Brain-Computer Interface Future Industry Cluster Area, located in Xinhongqiao, Minhang District.International MedicineCentrally located, adjacent to the Hongqiao Branch of Huashan Hospital Affiliated with Fudan University. Leveraging the clinical resource advantages of Huashan Hospital's neurosurgery department, the Shanghai Science and Technology Commission and the Minhang District government are constructing specialized platforms in the cluster area for concept verification, testing and certification, pilot processing, etc., attracting and nurturing outstanding teams and enterprises from China and abroad to settle and develop, assisting Shanghai in becoming an innovation source and industry highland for brain-computer interface with global influence.

Paralyzed Individuals "Brain-Control" Wheelchairs and Robotic Dogs

Brain-computer interface technology is divided into two major categories: invasive and non-invasive. Invasive technology involves implanting signal recording devices, such as electrodes, into the cerebral cortex through neurosurgery to achieve high-throughput neural signal acquisition. Compared with invasive technology, non-invasive technology has lower risks but more limited functionality; its product form is typically wearable devices that attach electrodes for collecting EEG signals to the scalp. At the 2025 Brain-Computer Interface Conference, Shanghai-based companies unveiled several invasive products that have reached leading levels in China and internationally.

"How do you make the wheelchair turn?" "I want it to turn wherever I wish, and it does." A video released by StairMed showcased the "brain-control" effects achieved in clinical trials of a 64-channel brain-computer interface system. After being implanted with this invasive device, an elderly person could freely maneuver the wheelchair—circling around flower beds in the neighborhood. As Zhang Shiqiang, the Chief Product Officer of the company, put it: "Nothing special, just brain proficiency." Paralyzed patients implanted with this device can also "brain-control" robotic dogs, making them carry purchased items from outdoors into the bedroom.

Elderly Individuals Can Freely "Mind-Control" Wheelchairs After Implantation of 64-Channel Brain-Computer Interface System.

According to reports, StairMed's newly developed product – a 256-channel brain-computer interface system – is about to enter clinical trials. Compared with the 64-channel system, it can collect more EEG signals, further enhancing the ability of paralyzed patients to control external devices. The company predicts that after patients are implanted with this product, their "brain-controlled" speed for phones and computers will approach that of normal individuals, enabling smooth control of five-axis robotic arms and preliminary control of embodied intelligence.Robot. When integrated with specialized artificial intelligence models, it can also help aphasia patients reconstruct language function.

StairMed was founded by two scientists from the Center of Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences. Leveraging ultra-flexible electrode technology, the company has developed electrodes with dimensions approximately 1/100 the thickness of a hair, achieving international leadership in terms of flexibility, size, biocompatibility, and channel density.

Another product released by the company is the Deep Brain Neurostimulation System. Unlike the "brain-control" devices implanted in paralyzed patients, the deep brain neurostimulation system is a "control-brain" device that primarily stimulates electrically after implantation and secondarily collects EEG signals. It targets chronic and treatment-resistant diseases such as Parkinson's disease, epilepsy, and refractory depression, and has completed preliminary functional verification in rats, monkeys, and humans.

Promote the Clinical Transformation of Blindness Reversal Technology

"Our company is preparing to settle in the future industrial cluster area for brain-computer interfaces." Liu Bing, founder of MingVision Brain-Computer and associate researcher at the Institute of Automation, Chinese Academy of Sciences, told reporters. At the new product launch event, MingVision Brain-Computer showcased a visual reconstruction demonstration system. The raw images captured by the camera are decoded into black-and-white contour images of various objects, maintaining a one-to-one pixel correspondence with the original images. Liu Bing explained that the black-and-white contour images represent the world as seen by blind individuals after receiving brain-computer interface implants. "Our goal is to enable completely blind individuals to achieve vision equivalent to 0.1."

Dr. Liu Bing Speaks at the 2025 Brain-Computer Interface Conference.

On the track of brain-computer interfaces enabling the blind to "regain sight," Neuralink, founded by Elon Musk, has made a breakthrough. The company's "Blindsight" device has completed animal experiments and has been designated as a "breakthrough device" by the U.S. Food and Drug Administration (FDA). "Blindsight" does not follow the ocular repair route but directly targets the visual cortex of the brain, stimulating the cortex to generate visual perception in the blind.

The research and development concept of MingShi Brain-Computer Interface is consistent with Neuralink, bypassing ocular tissues to output neuro-electrophysiological signals to the brain's visual cortex. Recently, this company announced the world’s first functional interaction verification for the reconstruction of "complex graphics + multiple colors" in vision. Clinical studies conducted at Wuhan Tongji Hospital have shown that intracranial electroencephalogram recording and electrical stimulation of the occipital visual cortex allowed patients' visual perception to evolve from simple light spots to complex shapes, also enabling them to perceive colors.

Speaking of this breakthrough, Liu Bing said, "We have proven that it is feasible to reconstruct meaningful visual information in the visual cortex through a brain-computer interface. The next step is to fully promote the clinical transformation of the technology." After settling in the Future Industry Cluster Zone, MingSight Brain Interface is expected to collaborate with the Department of Neurosurgery at Huashan Hospital to advance clinical trials. Plans are also in place to cooperate with the Shanghai Medical Device Inspection and Research Institute next year to carry out type inspections for medical devices under development.

This company will also enjoy the special policies of the cluster area. According to reports, the brain-computer interface special policy formulated by Minhang District focuses on cultivating industrial clusters, supporting technological and product innovation efforts, and providing full-cycle ecological support in three aspects, offering enterprises comprehensive "substantial" support to create a favorable ecosystem of "leading by the head and gathering startups."

EEG Large Model Upgrades Non-Invasive Technology

On the non-invasive route, the "Brain-Computer Interface System Based on Large-Scale EEG Models" developed by Yansi Brain-Inspired significantly improved the accuracy of "brain control." At a Chinese chess tournament held last month, participants wore Yansi Brain-Inspired's non-invasive EEG caps and used their thoughts to complete each move.

The EEG signals collected by the EEG cap have a low signal-to-noise ratio and are susceptible to interference. How can precise control be achieved? Li Meng, chief scientist of NeuroInspired and researcher at the Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, said that large-scale EEG models can enhance the decoding ability of brain-computer interfaces. This company, located in the Future Industry Cluster Zone, has collaborated with Huashan Hospital to combine big data from intracranial EEG signals with the algorithmic advantages of their research team. They have developed an EEG large model with 5 billion parameters, setting a world record in the field of brain-computer interfaces.

Li Meng believes that, with the continuous improvement in the performance of non-invasive technologies, non-invasive brain-computer interfaces will provide entirely new interaction methods for trillion-dollar markets such as consumer electronics, smart homes, and the metaverse, and will also be promoted and applied in the rehabilitation medical industry.

Today's brain-computer interface competition showcased the current level of this futuristic human-machine interaction method. This national competition featured four tracks: emotion perception, fatigue monitoring, brain-controlled robotic arms, and brain-controlled racing cars. At the brain-controlled robotic arm and brain-controlled racing car venues, reporters observed participants from various universities, research institutes, and enterprises wearing EEG caps, concentrating intently to "control" robotic arms and race cars with their minds, competing in terms of accuracy and speed.

Researchers wearing EEG caps participate in a brain-controlled car race.

Zheng Siyu, a contestant from Shaonao Technology, told reporters that the brain-controlled robotic arm competition required each team to "brain-control" the robotic arm to pick up wooden blocks of four colors on the table and place them into four trash bins of corresponding colors. "This competition has three key points: the brain electrolysis decoding ability supported by algorithms, the path design level of the robotic arm, and the anti-interference capability of the brainwave acquisition equipment." It was reported that the brainwave acquisition rehabilitation training device developed by Shaonao Technology obtained the first medical device registration certificate for brain-computer interface in Shanghai last year. The device has been introduced into nearly 30 hospitals across China and is being used for rehabilitation training of stroke patients, helping them rebuild their nervous systems.

"Currently, Shanghai has achieved a series of phased breakthroughs in the field of brain-computer interface, giving birth to several milestone results," said Qu Wei, Deputy Director of the Municipal Science and Technology Commission. The world's first implantable brain-computer interface GCP (Good Clinical Practice) clinical trial, China's first invasive brain-computer interface system registration prospective clinical trial, the world's first real-time Chinese decoding IIT trial (investigator-initiated trial), and China's first brain-spine interface IIT trial have been successively carried out, with three invasive products entering the national stage.Innovative HealthcareSpecial Review Procedure for Devices. Meanwhile, various non-invasive products have been applied in motor function rehabilitation for stroke patients, screening for diseases such as depression and anxiety, and aiding sleep disorders, demonstrating promising potential.

Original Title: "Blind People See Shapes and Colors, Paralyzed Elderly ‘Brain-Control’ Wheelchairs — Shanghai Brain-Computer Interface Conference Brings Technological Dawn"

Column Editor: Huang Haihua

Source: Author: Liberation Daily, Yu Taoran