The National Medical Products Administration has approved YY/T 2029—2026, 'Brain-Computer Interface Medical Devices — Quality Requirements and Evaluation Methods for AI-Algorithm EEG Datasets,' the world's first recommended industry standard specifically targeting AI-algorithm electroencephalogram (EEG) datasets for brain-computer interfaces.
The standard, which takes effect on September 1, 2027, systematically specifies quality requirements across the full lifecycle of EEG data — collection, processing, labeling, storage and access — and was drafted with participation from institutions including Tianjin University's Brain-Computer Haihe Laboratory.
It is China's third brain-computer-interface medical-device standard and the first globally to focus on AI algorithms specifically. By setting a unified yardstick for datasets that AI decoding depends on, it resolves the industry's long-standing problem of 'everyone speaking their own language' on data.
The standard particularly benefits the non-invasive BCI segment, providing hard references for enterprise R&D and registration review, and clearing the channel for innovative products to enter the market — marking the formal approach of non-invasive brain-computer interfaces to standardized, commercial deployment.
Industry observers say 'standard first, clinical follow' is the right sequence for a frontier field: it gives developers certainty, shortens the path from lab to clinic, and helps China stake an early lead in a technology broadly seen as strategic for healthcare and human-machine interaction.
