Scholarlink

Here we present a deep learning approach that paves the way for minimally-invasive, optically-based Brain-Computer Interfaces. We used a recurrent encoder-decoder network (LSTM-encdec) to accurately decode limb trajectories of a running mouse from two-photon calcium imaging. The LSTM-encdec takes neural signals recorded with two-photon calcium imaging (7.8 Hz) as inputs and outputs a longer sequence of limb coordinates (30 Hz), accurately predicting trajectories of an ipsilateral limb from neural activity within the sensorimotor cortex in a single cortical hemisphere. These results present a significant advance in the field of BCI by expanding the realm of usable technology for recording movement-related neural activity.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Toward an Optical BCI: Overcoming the Limitation of Low Sampling Rate for Decoding Limb Movements

  • 章节

  • 发表日期：

    2025年

查看全文
  • Seungbin Park,
  • Megan Lipton,
  • Maria Dadarlat

摘要

Here we present a deep learning approach that paves the way for minimally-invasive, optically-based Brain-Computer Interfaces. We used a recurrent encoder-decoder network (LSTM-encdec) to accurately decode limb trajectories of a running mouse from two-photon calcium imaging. The LSTM-encdec takes neural signals recorded with two-photon calcium imaging (7.8 Hz) as inputs and outputs a longer sequence of limb coordinates (30 Hz), accurately predicting trajectories of an ipsilateral limb from neural activity within the sensorimotor cortex in a single cortical hemisphere. These results present a significant advance in the field of BCI by expanding the realm of usable technology for recording movement-related neural activity.