https://doi.org/10.1029/2026GL122914

China has approved and issued the country’s third medical-device standard for brain-computer interfaces (BCIs), which is also the world’s first such standard for BCI medical devices using artificial intelligence (AI) to process electroencephalogram (EEG) data.

BCI technology establishes a direct communication route between the brains of humans or animals and external devices. By capturing and decoding neural signals, BCI enables seamless interaction between brain and machine, ultimately allowing control over assistive devices to execute intended actions.

The standard, issued by the National Medical Products Administration, specifies quality requirements and evaluation methods for EEG datasets used by AI algorithms in BCI medical devices. It applies to datasets used in research and development, production, quality control and other stages of BCI medical device technology that uses AI algorithms to post-process EEG signals.

The standard covers requirements for the collection, processing, annotation, storage and access of EEG data, as well as methods for testing dataset quality. It will help address common industry challenges, including inconsistent quality-control criteria and non-standardized dataset development.

China previously issued two BCI medical-device standards in 2025, covering terminology and test methods for the sensing and response performance of implantable neural stimulators with closed-loop functions.