A method to determine the interdependent relationships between biomechanical variables in artificial neural network models: the case of lower extremity muscle activity and body sway
A method to determine the interdependent relationships between biomechanical variables in artificial neural network models: the case of lower extremity muscle activity and body sway
复制标题
一种确定人工神经网络模型中生物力学变量之间相互依赖关系的方法：以下肢肌肉活动和身体摇摆为例
DOI：
10.1016/j.neucom.2003.05.002
复制
发表时间：
2004
期刊：
影响因子：
6
通讯作者：
L. Haugh
中科院分区：
文献类型：
--
作者：
Gongbing Shan;Ge Wu;L. Haugh
本研究探讨了使用人工神经网络权值来量化投入产出关系的可能性。我们推导了一个一阶系数来计算第i个输入对第k个输出的相对贡献，称为Qki。理论分析和数值模拟表明，qkk依赖于输入/输出函数的幅度、频率和相移，从而可以识别出对输出贡献最大的输入。这种方法被应用于研究肌肉活动和身体摇摆在人类安静的立场之间的关系。结果证明了它在研究人体姿势控制问题上的潜力。
This study examined the possibility of using artificial neural network weights to quantify input–output relationship. We derived a first-order coefficient to calculate the relative contribution from the ith input to the kth output, called Qki. Theoretical analysis and numerical simulation suggested that Qkidepended on the magnitude, frequency and phase shift of the input/output function, making it possible to identify those inputs that have the greatest contribution to the output. This approach was applied to investigate the relationship between muscle activities and body sway during human quiet stance. The results demonstrated its potential for studying human postural control issues.