跳到正文
原文
Nature Reviews Neurology· Yu Qi·· 2026-09-14AI 评分29

高性能手写脑机接口:Nature Reviews Neuroscience 期刊俱乐部评述

High-performance handwriting brain–computer interfaces

AI 导读

Nature Reviews Neuroscience 发表评述,回顾Willett等人2021年报道的手写脑机接口:从四肢瘫痪参与者运动皮层实时解码手写意图,打字速度达每分钟90字符、准确率94%。评述指出手写比点对点伸取动作具有更高的时间维度,字符间神经距离更大、更抗噪,并已启发语音解码等复杂序列行为研究。

正文
  • Journal Club
  • Published: 14 September 2026

Brain–computer interfaces

Nature Reviews Neuroscience (2026) Cite this article

Brain–computer interfaces (BCIs) can convert brain signals into instructions for external devices, which offers a potential route to restore communication for people with paralysis. For years, BCIs had advanced incrementally, with intracortical point-and-click typing plateauing at roughly 40 correct characters per minute — functional in principle, but far from natural communication. Then, in 2021, Willett and colleagues reported a BCI that decoded attempted handwriting from motor cortex activity in real time, achieving surprisingly high typing speeds of 90 characters per minute with 94% accuracy. Indeed, this BCI had approached the typing speed of able-bodied smartphone users, and, for many in the field, it was the moment when practical, high-performance neural communication truly felt within reach.

What made the work feel so elegant, beyond its striking performance, was the core insight behind its success. Rather than asking users to produce simple point-to-point reaching movements, the team leaned into handwriting, a motor skill honed over decades of practice. They recorded neural activity from the motor cortex while a participant with tetraplegia attempted to handwrite letters and symbols, and used a recurrent neural network to classify the characters from neural signals in real time. They demonstrated rigorously that handwriting carries far higher temporal dimensionality than point-to-point reaching movements, and that this richer spatiotemporal structure widens the neural distances between characters, which makes them inherently more robust to noise and easier to distinguish. It was a beautiful reminder that the brain’s own well-trained motor programmes could be an ideal substrate for BCIs. It shifted the field’s perspective on what kinds of movements to decode. The approach has since inspired a wave of research exploring other complex sequential behaviours, from speech decoding to internal thought translation.

This is a preview of subscription content, access via your institution

Access options

Access through your institution

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

27,99 € / 30 days

cancel any time

Subscription info for Chinese customers

We have a dedicated website for our Chinese customers. Please go to naturechina.com to subscribe to this journal.

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

39,95 €

Prices may be subject to local taxes which are calculated during checkout

Additional access options:

Author information

Authors and Affiliations

  1. MOE Frontier Science Center for Brain Science and Brain-machine Integration, Zhejiang University, Hangzhou, China

    Yu Qi

Authors

  1. Yu Qi

Corresponding author

Correspondence to Yu Qi.

Ethics declarations

Competing interests

The author declares no competing interests.

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Qi, Y. High-performance handwriting brain–computer interfaces. Nat. Rev. Neurosci. (2026). https://doi.org/10.1038/s41583-026-01082-w

Download citation

  • Published: 14 September 2026

  • Version of record: 14 September 2026

  • DOI: https://doi.org/10.1038/s41583-026-01082-w

来源:Nature Reviews Neurology · nature.com