Research progress on multiscale entropy algorithm and its application in neural signal analysis

脑神经电生理信号的内在特征变化能够反映脑功能的正常与否，因此有效的特征提取分析方法有利于脑功能异常的早期诊断与相关疾病的治疗。近年来的研究表明，神经电信号具有非线性和多尺度的特性。基于此，科研人员近来发展了适用于多尺度非线性信号分析的多尺度熵（MSE）算法，并在神经信息科学领域得到了广泛应用。本文对 MSE 算法的原理和特性进行了介绍，并进一步介绍了几种在实际应用中针对 MSE 算法的一些不足而提出的相关改进算法。然后，对 MSE 及其改进算法在疾病诊断、脑功能分析以及脑-机接口等方面的应用进行了综述。最后，对上述各算法在神经信号分析中面临的挑战及其可能的发展方向进行了探讨。.; Changes in the intrinsic characteristics of brain neural activities can reflect the normality of brain functions. Therefore, reliable and effective signal feature analysis methods play an important role in brain dysfunction and relative diseases early stage diagnosis. Recently, studies have shown that neural signals have nonlinear and multi-scale characteristics. Based on this, researchers have developed the multi-scale entropy (MSE) algorithm, which is considered more effective when analyzing multi-scale nonlinear signals, and is generally used in neuroinformatics. The principles and characteristics of MSE and several improved algorithms base on disadvantages of MSE were introduced in the article. Then, the applications of the MSE algorithm in disease diagnosis, brain function analysis and brain-computer interface were introduced. Finally, the challenges of these algorithms in neural signal analysis will face to and the possible further investigation interests were discussed.