浙江大学提出 Brain-Writing-BERT 手写脑机接口框架，汉字解码在线准确率达 94.45%

Article
Open access
Published: 05 November 2026

Xiaomeng Yang^1,2,3^na1,
Jie Qin^1,2,3^na1,
Xinzhu Xiong^1,2,3,
Junming Zhu⁴,
Jianmin Zhang ORCID: orcid.org/0000-0002-3184-1502⁴,
Yueming Wang ORCID: orcid.org/0000-0001-7742-0722³ &
…
Yu Qi ORCID: orcid.org/0000-0002-9789-8907^1,2,3,5

Communications Biology (2026) Cite this article

We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.

Abstract

Handwriting Brain-Computer Interfaces (BCIs) offer a direct brain-to-text pathway, holding significant promise for neural rehabilitation. However, existing research has primarily focused on alphabetic languages. Unlike the sequential nature of alphabetic handwriting, logographic characters like Chinese incorporate complex spatial structures that embed semantic information, presenting a greater challenge for neural decoding. To address this, we propose Brain-Writing-BERT, a novel BCI framework that learns the structural semantic priors from logographic characters to facilitate decoding. We evaluate our system with a clinical participant in Chinese character decoding tasks. All clinical and experimental procedures were registered in the Chinese Clinical Trial Registry (ref.ChiCTR2100050705, date of registration: 03/09/2021, registry name: Research on Brain-Computer Interfaces for Individuals with High Cervical Spinal Paraplegia). The proposed BCI achieves an online accuracy of 94.45% for a 295-character vocabulary, with an average speed of 8.55 ± 2.77 characters per minute. This study advances high-performance handwriting BCIs for logographic systems, broadening communication possibilities for diverse languages.

Funding

This work was partly supported by the Brain Science and Brain-like Intelligence Technology-National Science and Technology Major Project (2025ZD0218800), the Key R&D Program of Zhejiang (2023C03001), the Zhejiang Provincial Natural Science Foundation of China (LR24F020002), the National Natural Science Foundation of China (62336007), and the Starry Night Science Fund of Zhejiang University Shanghai Institute for Advanced Study (SN-ZJU-SIAS-002). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Author information

Author notes

These authors contributed equally: Xiaomeng Yang, Jie Qin.

Authors and Affiliations

MOE Frontier Science Center for Brain Science and Brain-Machine Integration, Zhejiang University, Hangzhou, China
Xiaomeng Yang, Jie Qin, Xinzhu Xiong & Yu Qi
Affiliated Mental Health Center & Hangzhou Seventh People’s Hospital, Zhejiang University School of Medicine, Hangzhou, China
Xiaomeng Yang, Jie Qin, Xinzhu Xiong & Yu Qi
College of Computer Science and Technology, Zhejiang University, Hangzhou, China
Xiaomeng Yang, Jie Qin, Xinzhu Xiong, Yueming Wang & Yu Qi
Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, China
Junming Zhu & Jianmin Zhang
State Key Laboratory of Brain-Machine Intelligence, Hangzhou, China
Yu Qi

Authors

Xiaomeng Yang
Jie Qin
Xinzhu Xiong
Junming Zhu
Jianmin Zhang
Yueming Wang
Yu Qi

Corresponding author

Correspondence to Yu Qi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Yang, X., Qin, J., Xiong, X. et al. Brain-writing-BERT for generalizable handwriting brain-computer interface. Commun Biol (2026). https://doi.org/10.1038/s42003-026-11194-9

Download citation

Received: 08 January 2026
Accepted: 06 October 2026
Published: 05 November 2026
DOI: https://doi.org/10.1038/s42003-026-11194-9