IEEE J Biomed Health Inform. 2015 Mar;19(2):406-17. doi: 10.1109/JBHI.2014.2316287. Epub 2014 Apr 9.
The impedance plethysmography (IP) has long been used to monitor respiration. The IP signal is also suitable for portable monitoring of respiration due to its simplicity. However, this signal is very susceptible to motion artifact (MA). As a result, MA reduction is an indispensable part of portable acquisition of the IP signal. Often, the amplitude of the MA is much larger than the amplitude of the respiratory component in the IP signal. This study proposes a novel filtering method to remove the high-amplitude MA's from the IP signal. The proposed method combines the idea of ε-tube loss function and an autoregressive exogenous model to estimate the MA while leaving the periodic respiratory component of the IP signal intact. Also, a regularization method is used to find the best filter coefficients that maximize the regularity of the output signal. The results indicate that the proposed method can effectively remove the MA, outperforming the popular MA reduction methods. Several different performance measures are used for the comparison and the differences are found to be statistically significant.
容积描记法（IP）长期以来一直用于监测呼吸。由于其简单性，IP 信号也适合用于呼吸的便携式监测。然而，该信号非常容易受到运动伪影（MA）的影响。因此，减少 MA 是便携式获取 IP 信号不可或缺的一部分。通常，MA 的幅度远大于 IP 信号中呼吸分量的幅度。本研究提出了一种新颖的滤波方法，可从 IP 信号中去除高幅度 MA。该方法结合了ε-管损失函数和自回归外生模型的思想，在保留 IP 信号中周期性呼吸分量的同时估计 MA。此外，还使用正则化方法来找到最佳滤波器系数，以使输出信号的正则性最大化。结果表明，该方法可以有效地去除 MA，优于流行的 MA 减少方法。使用了几种不同的性能指标进行比较，发现差异具有统计学意义。
