Recent successes demonstrate the expansion of the application scope of non-invasive brain-computer interfaces (BCIs) beyond electroencephalography. On August 26, Wuhan University Hospital reported the world's first clinical application of a high-resolution semi-invasive retinal BCI. Earlier still, Tianjin University and United Imaging introduced the 'uMR Shen Guan' solution—a comprehensive magnetic resonance BCI combining decoding, hardware, and visualization.
This field is divided into three directions: invasive, semi-invasive, and non-invasive. Systems that do not require surgery keep sensors outside the skull, sacrificing some spatial accuracy for the possibility of reusability and portability. Within the non-invasive approach, traditionally dominated by EEG, functional ultrasound and other signal acquisition methods are attracting scientific interest as the industry strives to improve spatial resolution, multimodal decoding, and closed-loop feedback.
Commercial strategies are developing in various directions. The most obvious is medical rehabilitation: due to clearly defined indications, device registration, and demand from hospitals, this area represents the easiest path for business. Neuracle has developed EEG-VR rehabilitation systems and VR rehabilitation tools using motor imagery, serving hospitals such as Huashan, Xuanwu, and the China Rehabilitation Research Center. Its prospectus submitted to the Shanghai Stock Exchange indicates over 20 medical device registrations covering EEG apparatus, EMG and evoked potential tools, as well as transcranial electrical stimulation devices, including tES-EEG and TMS-EEG combinations. In 2025, the company ranked first nationally in terms of EEG equipment shipment volume and value.
The brain state and sleep monitoring segment targets a broader consumer audience with high usage frequency, but it still needs to prove long-term effectiveness and user retention. MindMatrices creates a portable polysomnography device, a deep brain stimulation system with temporal interference, and a closed sleep intervention system based on sleep rhythms and neural modulation.
Human-computer interaction holds the greatest potential, but it remains the furthest from stable revenue because markets are already dominated by keyboards, touchscreens, voice input, and eye tracking. Promising near-term niches include assisting people with severe paralysis and use as an input level in industrial settings where hands are occupied, while augmented and virtual reality, as well as embodied AI, are considered long-term additions.
Scaling will depend less on technological concepts than on engineering stability, clinical evidence, regulatory approvals, user retention, and data management. According to industry observers, the most likely first major market for non-invasive BCIs is medical rehabilitation, as the path from registration to hospital procurement is already visible.