China’s brain-computer interface technologies advance rapidly from laboratory to clinical application

Brain-computer interfaces form a technical system that creates a direct information pathway between the human brain and external equipment. Such systems capture and decode neural activity, translating users’ motor, linguistic or sensory intentions into control commands for devices. Certain variants can also deliver electrical stimulation or sensory feedback to the nervous system, forming a closed two-way loop.

Three distinct technical routes for brain-computer interfaces are being explored in parallel across China, pushing this technology steadily out of laboratories and into clinical settings. These categories are non-invasive systems with electrodes placed on the scalp, semi-invasive systems where electrodes sit outside brain tissue on or beneath the dura mater, and fully invasive systems with electrodes implanted inside the brain.

The Haihe Laboratory of Brain-Computer Interaction and Human-Machine Integration, built around Tianjin University in the Binhai High-tech Area of Tianjin, has developed more than 30 medical devices under its Shenggong product range through sustained research. Non-invasive brain-computer interfaces collect electroencephalogram signals directly on the scalp via tools such as electroencephalogram caps. These systems carry stronger safety profiles and serve wider groups of users across more scenarios. Products from the Shenggong series have been deployed in rehabilitation medicine, psychiatry, neurocritical care and audio-visual diagnosis, benefiting thousands of patients. The Shenggong-Shen’er auditory assessment system extracts brain waveforms to read neural responses to sound, enabling clinicians to fine-tune cochlear implants and hearing aids so patients can clearly perceive sound rather than merely detect it. The Shenggong-Shengao hydrocephalus diagnostic platform cuts diagnostic times from two to three days down to 30 minutes and substantially lowers misdiagnosis rates.

A semi-invasive brain-computer interface known as Beinao No.1 has delivered notable clinical outcomes. A patient with severe T12–L1 spinal trauma and complete loss of lower-limb mobility received synchronous implantation of Beinao No.1 and a sequential spinal cord stimulation system. After one year of rehabilitation training, the patient can walk independently with brace support and regain voluntary control over bladder and bowel functions. Beinao No.1’s original iteration uses a 128-channel fully implantable flexible wireless electrode array positioned outside the dura mater, balancing safety and therapeutic performance. Multi-centre registration clinical trials for Beinao No.1 began in March 2025, covering indications including spinal cord injury, hemiplegia caused by stroke, motor impairment of limbs and aphasia triggered by amyotrophic lateral sclerosis. Patient recruitment for these trials has finished, with an application for medical device registration planned for 2027.

Beinao No.2, an invasive brain-computer interface, is progressing alongside this work. It features a 512-channel wireless fully implanted flexible intracortical electrode array, developed to match international multi-thousand-channel products. The system targets patients with severe paralysis requiring highly precise motor decoding, and human clinical validation is scheduled to start within the year.

In May this year, Beijing Zhiran Medical Technology Co., Ltd launched registration clinical trials for China’s first over-100-channel invasive brain-computer interface system, marking a new phase of clinical translation for high-flux invasive brain-computer technology. The 128-channel system consists of flexible electrodes and a signal collector. Its ultra-thin biocompatible flexible material reduces immune reactions after implantation and supports real-time high-precision decoding of fine brain electrical signals. The fully buried wireless signal collector supports wireless charging. The clinical programme aims to help tetraplegic patients with spinal cord injury operate brain-controlled cursors and restore partial hand motor function. The trials are expected to conclude in 2027.

Additional clinical trials for brain-computer interfaces at varying development stages are underway in Shanghai, Guangzhou, Chengdu, Shenzhen and Hefei. Most high-flux implantable brain-computer interfaces remain at the early clinical validation stage. Some observed functional improvements stem from combined intervention involving brain-computer interfaces, neural stimulation and long-term rehabilitation training. Larger sample sizes, controlled studies and extended follow-up will confirm therapeutic durability, suitable patient groups and the independent contribution of each technical component.

Clinical diagnosis and treatment using brain-computer interfaces in China has entered a critical phase of clinical verification and registration trials. The country holds clear comparative advantages, including a large clinical patient pool affected by spinal cord injury, stroke, Parkinson’s disease and epilepsy. Collaboration between neurosurgery, rehabilitation medicine, engineering teams and industrial players continues to deepen, with clinicians deeply engaged in technical development and clinical protocol design to align technical targets with genuine patient needs. Brain-computer interfaces can only resolve a subset of conditions beyond conventional medical capabilities, and functional reconstruction cannot happen instantly. Safety, efficacy and stability of these technologies await further testing through prolonged practice.

Further functional expansion is anticipated in the coming years. Within the next five years, the technology may deliver motor function substitution, neural rehabilitation and speech reconstruction to improve self-care capacity and quality of life for patients. Over five to ten years, brain-computer interfaces may unlock more advanced capabilities. Two-way integrated brain-computer systems will evolve from simple neural signal reading to writable neural modulation, triggering stimulation of native muscles or spinal nerves after algorithms decode user intent. Simultaneous multi-region brain decoding will expand coverage from the primary motor cortex to visual, emotional and cognitive brain regions, greatly broadening the scope of decoded information. Beyond functional repair and tactile reconstruction, the technology may also support intervention for mental health disorders, offering new treatment pathways for patients living with depression.

The scope of brain-computer interfaces stretches far beyond medical diagnosis and treatment. Listed as a future industry under China’s 15th Five-Year Plan outline, brain-computer interfaces will drive domestic scientific progress and industrial upgrading. The technology has shown promising application prospects in aerospace, education, entertainment and transport, with even wider fields of use set to emerge in the future. In the foreseeable future, brain-computer interfaces may enable deep collaboration between human brains, artificial intelligence and robots, and extend to cognitive and sensory enhancement for healthy individuals. These concepts still face technical hurdles alongside ethical considerations covering neural privacy, personal autonomy, equitable access and risks associated with non-therapeutic implantation.

As a next-generation human-machine interaction platform, its core strength lies in high-bandwidth two-way information exchange between humans and machines. Human speech and typing transfer merely dozens of bytes per second, while exchanges between intelligent systems can run at hundreds of megabytes or several gigabytes each second. Continuous iterative upgrades of brain-computer hardware, paired with advances in wireless communications and artificial intelligence, will support higher-bandwidth bidirectional neural data transmission, opening extensive possibilities for social transformation.

Building this industrial ecosystem requires coordinated efforts across government, academia, hospitals and enterprises. Core technical challenges span electrode materials, micro-nano fabrication, dedicated mixed-signal chips, integrated wireless systems and real-time decoding algorithms. Industrial participants need to focus on underlying technologies and system integration rather than assembling imported core components, to deliver independently controllable, high-performance solutions.

Standardisation and policy frameworks also require refinement. Unified criteria for trial enrolment and outcome evaluation must be established to raise the quality of clinical research. Standardised multi-modal neuroelectrophysiological databases will enable harmonised clinical data and cross-institutional sharing. Mechanisms covering product costing, hospital admission and medical insurance reimbursement will be refined to support safe, effective, replicable and affordable clinical deployment.

The interdisciplinary nature of brain-computer interfaces combines neuroscience, electronic engineering, materials science, chip design and algorithms. Clinical deployment demands close teamwork from medical and rehabilitation professionals, creating widespread gaps in skilled personnel. Tianjin University has launched China’s first undergraduate degree programme in Brain-Computer Science and Technology, accepting applicants nationwide. Joint training programmes between universities, hospitals and enterprises will expand the talent pipeline.

Breakthrough disruptive technologies grow from original discoveries in basic research, and the depth of fundamental neuroscience research defines the ceiling for innovation. While China has made rapid progress in neuroscience and brain cognition, original discoveries remain limited. Sustained investment in biomedical, neuroscience and brain science research will encourage pioneering fundamental work.

China’s brain-computer interface sector has entered a robust growth phase. Coordinated, practical work across government, academia, hospitals and industry will unlock the full potential of this emerging technology.