Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

The Association Between Inclusive Leadership and Job Performance in Nurses: Exploring the Mediation Roles of Grit and Work Engagement Based on a Cross-Sectional Study.

Journal of nursing management·2026
Same author

Inclusive Leadership Is Associated With Nurse-Perceived Nurse-Patient Relationship Quality Through the Mediation Link of Self-Control and Resilience: A Cross-Sectional Study.

Journal of nursing management·2026
Same author

MicroRNA expression profiling and functional analysis of CDH3 during oogenesis in the Chinese alligator (<i>Alligator sinensis</i>).

Current zoology·2026
Same author

The Quality of Grape Berries and Wine Is Enhanced Due to the Intercropping of Green Manure by Regulating Soil Microecology.

Foods (Basel, Switzerland)·2026
Same author

Adjunctive acupuncture for sepsis-associated acute gastrointestinal injury: a systematic review, meta-analysis, and exploratory Bayesian network meta-analysis.

Frontiers in medicine·2026
Same author

Treatment of cervical squamous cell carcinoma with transabdominal lymph node resection and QM-C2 surgery: a surgical technique.

Gynecology and pelvic medicine·2026

相关实验视频

Updated: Sep 16, 2025

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding
11:25

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Published on: July 26, 2013

43.5K

评估卷积神经网络架构用于神经驱动器从高密度表面电肌图解码.

Jirui Fu, Helen J Huang, Yue Wen

    IEEE ... International Conference on Rehabilitation Robotics : [proceedings]
    |July 11, 2025
    PubMed
    概括
    此摘要是机器生成的。

    这项研究比较了1D和3DCNN模型来解码来自HD-sEMG信号的神经驱动器. 1D CNN具有更大的窗口,而3D CNN则提供较低的延迟但更高的计算成本.

    更多相关视频

    Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
    08:15

    Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision

    Published on: March 28, 2025

    751
    Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks
    11:18

    Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

    Published on: March 2, 2015

    10.4K

    相关实验视频

    Last Updated: Sep 16, 2025

    Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding
    11:25

    Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

    Published on: July 26, 2013

    43.5K
    Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
    08:15

    Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision

    Published on: March 28, 2025

    751
    Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks
    11:18

    Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

    Published on: March 2, 2015

    10.4K

    科学领域:

    • 生物医学工程 生物医学工程
    • 神经科学是一个神经科学.
    • 机器学习 机器学习

    背景情况:

    • 卷积神经网络 (CNN) 模型,包括1D和3D CNN,在解码来自高密度表面电肌图 (HD-sEMG) 信号的神经驱动器方面表现有前途.
    • 使用相同数据集进行神经驱动解码的1D与3DCNN的比较性能仍未得到充分探索.

    研究的目的:

    • 评估和比较1D CNN和3D CNN模型在提取神经驱动器作为累积尖峰列车 (CST) 的性能.
    • 调查关键参数,特别是窗口和步骤大小对两种CNN维度的解码精度的影响.
    • 评估与神经驱动器解码的1D和3DCNN模型相关的计算成本.

    主要方法:

    • 利用了三名参与者的胃肌肌肉的实验性HD-sEMG数据集.
    • 经过训练和验证的1D CNN和3D CNN模型将神经驱动器解码为累积尖峰列车 (CST).
    • 使用F1分数和相关系数与各种窗口和步骤大小的卷积内核补偿 (CKC) 算法进行模型性能比较.

    主要成果:

    • 1D CNN以更大的窗口大小 (80-120个样本) 实现了峰值性能 (F1得分:0.84,相关性:0.94).
    • 3D CNN在较小的窗口尺寸 (20-40个样本) 中达到峰值性能 (F1得分:0.83,相关性:0.92),这表明延迟较低.
    • 两种模型都显示了随着步骤大小的增加而降低了性能,3D CNN的计算要求明显更高 (938G FLOP与1D CNN的60G FLOP相比).

    结论:

    • 在基于HD-sEMG的神经驱动器解码的1D和3D CNN架构之间存在显著差异.
    • 最佳的参数选择 (窗口/步骤大小) 是至关重要的,并且在1D和3D CNN之间有所不同,以最大限度地提高准确性并最大限度地减少延迟.
    • 这些发现为精确和实时的神经驱动器解码应用选择适当的CNN模型和参数提供了指导.