Purpose

The purpose of this paper is to provide appropriate methods for reducing the abnormalities on the extracted fetal heart signal from the maternal electrocardiogram (ECG).

Design/methodology/approach

In this regard, the extracted signal of the fetal heart from the mother, improved using an active noise cancelation (ANC) system. It uses commonly adaptive algorithms of normalized least mean squares (NLMS). In the present paper, fetal extraction and denoising methodology are proposed. This methodology uses a combination of the NLMS algorithm with Savitzky–Golay (S-G) filter.

Findings

The obtained results show that a combination of NLMS algorithm with filter coefficient of 15 and µ = 0.02 and S-G filter has a better qSNR (qSNR = 3.6727) and good performance for fetal ECG extraction in comparison with the other works for average fmSNR in the range of −30 to −15 dB. Also, with considering the SNR value of −24.7 dB before filtering and SNR = 3.1861 dB after filtering; the SNR improvement of 27.8861 dB has been obtained.

Originality/value

A new method in the extract and noise reduction of fetal ECG from maternal ECG by the combination of NLMS algorithm and S-G filter is proposed.

中文翻译：

---

结合NLMS算法和SG滤波器对胎儿心脏信号进行恢复和噪声消除

目的

本文的目的是提供适当的方法，以减少从产妇心电图（ECG）提取的胎儿心脏信号中的异常。

设计/方法/方法

在这方面，使用主动噪声消除（ANC）系统改善了从母亲那里提取的胎儿心脏信号。它使用归一化最小均方（NLMS）的通用自适应算法。本文提出了胎儿的提取和去噪方法。该方法结合了NLMS算法和Savitzky-Golay（SG）滤波器的组合。

发现

获得的结果表明，与平均范围内的平均fmSNR的其他工作相比，结合滤波系数为15和µ = 0.02的NLMS算法和SG滤波器，具有更好的qSNR（qSNR = 3.6727）和良好的胎儿心电图提取性能。 -30至-15 dB。同样，考虑到滤波前的SNR值为-24.7 dB，滤波后的SNR = 3.1861 dB；SNR改善了27.8861 dB。

创意/价值

提出了一种结合NLMS算法和SG滤波器从母体心电图中提取和降低胎儿心电图的新方法。