期刊
TRENDS IN NEUROSCIENCES
卷 45, 期 3, 页码 176-183出版社
CELL PRESS
DOI: 10.1016/j.tins.2021.12.006
关键词
-
资金
- Engineering and Physical Sciences Research Council [EP/T020970/1]
- Wellcome Trust Senior Research Fellowship [215575/Z/19/Z]
This study argues that brain-computer interfaces (BCIs) for movement restoration need to better decode the user's intent. The researchers point out that the neural activity in the primary motor cortex provides both too little and too much information, which may affect the accuracy of decoding. To address this issue, the researchers suggest integrating additional information from multiple brain regions to enhance the interpretability of BCIs.
Brain-computer interfaces (BCIs) for movement restoration typically decode the user's intent from neural activity in their primary motor cortex (M1) and use this information to enable 'mental control' of an external device. Here, we argue that activity in M1 has both too little and too much information for optimal decoding: too little, in that many regions beyond it contribute unique motor outputs and have movement-related information that is absent or otherwise difficult to resolve from M1 activity; and too much, in that motor commands are tangled up with nonmotor processes such as attention and feedback processing, potentially hindering decoding. Both challenges might be circumvented, we argue, by integrating additional information from multiple brain regions to develop BCIs that will better interpret the user's intent.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据