The fetal electrocardiogram by independent component analysis and wavelets

Once the fetal electrocardiogram (FECG) waveforms from ECG on the maternal abdomen are detected, the fetal P wave and T wave cannot always be identified by using continuous wavelet transform (CWT). We took noninvasive FECG from the maternal abdomen, extracted it from the maternal electrocardiogram waveforms after an Independent Component Analysis (ICA), and identified the features of those waveforms by using CWT. We also simultaneously analyzed the observed signals by Primary Component Analysis (PCA). FECG has been extracted by ICA from 25 of 30 pregnant women. The fetal P wave and T wave could be identified in 21 of the 25 cases. FECG was extracted by PCA in only one case. ICA is superior to PCA, whose separation quality highly depends on the careful positioning of the electrodes. We believe that after ICA, FECG obtained by the wavelet theory based method will become a powerful tool for the differential diagnosis of fetal arrhythmias.     

一种基于单通道腹部信号的胎儿心电提取算法

设计一种基于单通道孕腹部信号的胎儿心电提取算法,分别提取出母亲心电和胎儿心电,并计算出母亲心率和胎儿心率。首先对单通道孕腹部信号进行k-TEO(k=19)变换,突出母亲心电的QRS波,从而通过简单的阈值法确定母亲心电的R波位置,接着通过在相邻R波间重采样以获得相同的R-R间期T,这样经过一个间隔为T的梳状滤波器就可以分离出相同R-R间期的母亲心电,然后再一次在相邻R波间进行重采样恢复原来的R-R间期就可以获得实际的母亲心电了。原始腹部信号减去上面提取的母亲心电后,胎儿心电QRS波的信噪比大大提高,通过再次应用提取母亲心电的算法即可得到“干净”的胎儿心电波形。选取Physionet数据库中的8 组(26 通道)孕腹部信号数据进行分析,计算每个通道数据的胎儿心电QRS波位置识别灵敏度、阳性检测率和准确性。结果表明,胎儿心电QRS波的识别准确率达到87.1%,其中有6 个通道达到100%。另外计算每个通道的母亲心率和胎儿心率并做统计分析,发现每一组中各个通道的母亲平均心率和胎儿平均心率都非常接近,同一组中各通道间母亲平均心率最大误差为0.1次/min, 而胎儿平均心率最大误差也只有0.9次/min,进一步证明算法的可靠性。     

Comparison of SVD methods to extract the foetal electrocardiogram from cutaneous electrode signals

The paper presents and compares three methods making use of the singular value decomposition (SVD) of a matrix to extract the foetal electrocardiogram (FECG) from cutaneously recorded electrode signals. The first method constructs a set of orthogonal foetal signals (the so-called principal foetal signals) from the recordings, but needs electrode positions far from the foetal heart, in addition to the abdominal electrodes that pick up a mixture of maternal and foetal electrocardiogram. An online adaptive algorithm has been developed such that a real-time implementation becomes feasible. The second method is a new online approach to a technique presented by van Oosterom. Although this method has some important drawbacks and is suboptimal as far as foetal signal-to-noise ratio is concerned. it is still very useful when only a foetal trigger is required, as the signal obtained is not a complete FECG, Finally, a third method is proposed, based on the generalised SVD and interpreted with the new concept of oriented signal-to-signal ratio. An online version is also presented for this method and some results are shown.     

Fetal heart rate monitoring based on independent component analysis

In this paper, an algorithm based on independent component analysis (ICA) for extracting the fetal heart rate (FHR) from maternal abdominal electrodes is presented. Three abdominal ECG channels are used to extract the FHR in three steps: first preprocessing procedures such as DC cancellation and low-pass filtering are applied to remove noise. Then the algorithm for multiple unknown source extraction (AMUSE) algorithm is fed to extract the sources from the observation signals include fetal ECG (FECG). Finally, FHR is extracted from FECG. The method is shown to be capable of completely revealing FECG R-peaks from observation leads even with a SNR=-200dB using semi-synthetic data.     

基于独立成分分析和量子粒子群算法的胎儿心电提取

胎儿心电图(FECG)是反映胎儿心脏电生理活动的一项客观指标,获取的FECG受到母体心电图(MECG)的干扰,如何快捷、有效的提取FECG成为重要的研究课题。在非侵入方式下,FECG的提取算法中独立成分分析(ICA)算法被认为是效果最好的方法,但现有求解其分解矩阵的算法收敛性能都不太高。量子粒子群(QPSO)算法是一种收敛于全局的智能优化算法。因此,提出了一种结合QPSO的ICA方法。研究结果表明,与其他在非侵入方式下的主要提取算法相比,这种方法能更清晰准确地提取出有用信号,为胎儿的健康检测提供了更好的方法。     

A Comparison of Single Channel Fetal ECG Extraction Methods

The abdominal electrocardiogram (ECG) provides a non-invasive method for monitoring the fetal cardiac activity in pregnant women. However, the temporal and frequency overlap between the fetal ECG (FECG), the maternal ECG (MECG) and noise results in a challenging source separation problem. This work seeks to compare temporal extraction methods for extracting the fetal signal and estimating fetal heart rate. A novel method for MECG cancelation using an echo state neural network (ESN) based filtering approach was compared with the least mean square (LMS), the recursive least square (RLS) adaptive filter and template subtraction (TS) techniques. Analysis was performed using real signals from two databases composing a total of 4 h 22 min of data from nine pregnant women with 37,452 reference fetal beats. The effects of preprocessing the signals was empirically evaluated. The results demonstrate that the ESN based algorithm performs best on the test data with an F1 measure of 90.2% as compared to the LMS (87.9%), RLS (88.2%) and the TS (89.3%) techniques. Results suggest that a higher baseline wander high pass cut-off frequency than traditionally used for FECG analysis significantly increases performance for all evaluated methods. Open source code for the benchmark methods are made available to allow comparison and reproducibility on the public domain data.     

两导心电信号独立分量分析算法的研究

根据独立分量分析(ICA)理论,要想在观测信号中提取出独立分量,观测信号的数目必须大于或等于独立分量的数目,因此要求采用ICA算法的胎儿心电图机导联数必须大于一定数目,但在实际应用中常常难以满足这个条件。故本文提出了一种基于少数导联心电(ECG)信号的胎儿心电(FECG)提取算法,结合FECG和自适应噪声抵消算法,从两导采集于孕妇腹部体表的ECG信号中提取FECG。实验表明,该方法能够获得清晰的FECG信号。     

Compact long-term recorder for the transabdominal foetal and maternal electrocardiogram

Foetal heart rate (FHR) monitoring is a proven means of assessing foetal health during the antenatal period. Currently, the only widely available instrumentation for producing these data is based on Doppler ultrasound, a technology that is unsuitable for long-term use. For nearly a century, it has been known that the foetal electrocardiogram (FECG) can be detected using electrodes placed on the maternal abdomen. Although these signals suggest an alternative means of FHR derivation, their use has been limited owing to problems of poor signal-to-noise ratio. However, the eminent suitability of the transabdominal FECG for long-term FHR monitoring has suggested that perseverance with the technique would be worthwhile. The paper describes the design, construction and use of a compact, long-term recorder of three channels of 24 h antenatal transabdominal data. Preliminary use of the recorder in around 400 short recording sessions demonstrates that FHR records of equivalent quality to those from Doppler ultrasound-based instruments can be extracted from such data. The success of FHR derivation is, on average, around 65% of the recording period from around 20 weeks gestation (although this figure is reduced from around 28–32 weeks, and the success rates exhibit a wide range when individual subjects are considered). These results demonstrate that the technique offers, not only a means of acquiring long-term FHR data that are problematic to obtain by other means, but also a more patient-friendly alternative to the Doppler ultrasound technique.     

基于快速定点独立分量分析算法的母胎心电信号分离

研究快速定点独立分量分析方法在母胎心电信号分离中的应用。采用此算法,在胎儿心电信号与母体心电信号可以视为相互独立的信号源的前提下,对来源于同一孕妇的观测信号进行独立分量分离。快速定点独立分量算法可以有效地分离出单个独立分量,得到的胎儿心电信号(FECG)较理想。采用独立分量分析方法,实现母胎心电信号分离,是一种值得尝试的信号处理方法。     

An application of reconfigurable technologies for non-invasive fetal heart rate extraction

This paper illustrates the use of a reconfigurable system for fetal electrocardiogram (FECG) estimation from mother's abdomen ECG measurements. The system is based on two different reconfigurable devices. Initially, a field-programmable analog array (FPAA) device implements the analog reconfigurable preprocessing for ECG signal acquisition. The signal processing chain continues onto a field-programmable gate array (FPGA) device, which contains all the communication and interfacing protocols along with specific digital signal processing blocks required for fundamental period extraction from FECG waveforms. The synergy between these devices provides the system the ability to change any necessary parameter during the acquisition process for enhancing the result. The use of a FPGA allows implementing different algorithms for FECG signal extraction, such as adaptive signal filtering. Preliminary works employ commercially available development platforms for test experiments, which suffice for the processing of real FECG signals from biomedical databases, as the presented results illustrate.     

Complete foetal ECG morphology recording by synchronised adaptive filtration

Present noninvasively measured indices of foetal stress are indirect and not sufficient. Foetal electrocardiography (FECG) is a potential noninvasive measurement of the foetus wellbeing which has been little utilised because of the difficulties of measuring it. The development of time-sequenced adaptive filters which are synchronised to the QRS complex by the use of Doppler echocardiography allowed the recording of relatively noise-free FECG. The paper describes the use of this technique for obtaining the complete complex of the FECG. Several sets of time-sequenced adaptive filters are combined to allow a multilead abdominal recording to produce a measurement system which rejects maternal ECG and enhances the FECG. Five subjects have been analysed, and their FECGs have been accurately reproduced with minimal changes of the filters' parameters.     

ECG-changes in the fetal lamb during asphyxia in relation to beta-adrenoceptor stimulation and blockade.

Progressive changes in the S-T interval of the fetal electrocardiogram (FECG) were studied in 14 lamb fetuses, acutely exteriorized and subjected to graded hypoxia. The aims of the study were to investigate whether beta-adrenoceptor stimulation and hypoxia exerted additive or potentiating effects on the FECG and several cardiovascular parameters and whether the hypoxic changes of the FECG could be blocked by beta-adrenoceptor blocking agents. The FECG changes were studied in order to correlate them with cardiovascular function, as measured by heart rate, mean arterial pressure, end diastolic pressure, maximum dP/dt and combined cardiac output, estimated by the thermodilution method, as well as with blood gases, acid base status, blood lactate and glucose. Injections of small doses (0.02 to 0.4 microg kg-1 min-1) of isoprenaline induced the same pattern of changes in the FECG as we have previously recorded during hypoxia. By increasing the isoprenaline dose an increase in the duration of the FECG changes and amplitude of the T-wave changes was obtained. Propranolol was found to completely abolish the FECG changes induced by isoprenaline, as well as by mild hypoxia. During severe hypoxia the FECG changes could not be abolished by propranolol. Our previous findings indicated that the hypoxic changes could be regarded as a sign of myocardial glycolysis. Thus, the present finding that even small doses of isoprenaline given to the fetus, initiates the same pattern of FECG changes corroborate this hypothesis.     

Research of fetal ECG extraction using wavelet analysis and adaptive filtering

Extracting clean fetal electrocardiogram (ECG) signals is very important in fetal monitoring. In this paper, we proposed a new method for fetal ECG extraction based on wavelet analysis, the least mean square (LMS) adaptive filtering algorithm, and the spatially selective noise filtration (SSNF) algorithm. First, abdominal signals and thoracic signals were processed by stationary wavelet transform (SWT), and the wavelet coefficients at each scale were obtained. For each scale, the detail coefficients were processed by the LMS algorithm. The coefficient of the abdominal signal was taken as the original input of the LMS adaptive filtering system, and the coefficient of the thoracic signal as the reference input. Then, correlations of the processed wavelet coefficients were computed. The threshold was set and noise components were removed with the SSNF algorithm. Finally, the processed wavelet coefficients were reconstructed by inverse SWT to obtain fetal ECG. Twenty cases of simulated data and 12 cases of clinical data were used. Experimental results showed that the proposed method outperforms the LMS algorithm: (1) it shows improvement in case of superposition R-peaks of fetal ECG and maternal ECG; (2) noise disturbance is eliminated by incorporating the SSNF algorithm and the extracted waveform is more stable; and (3) the performance is proven quantitatively by SNR calculation. The results indicated that the proposed algorithm can be used for extracting fetal ECG from abdominal signals.     

Reducing power line interference in digitised electromyogram recordings by spectrum interpolation

Interference from power lines (50 or 60 Hz) is the largest source of extraneous noise in many bio-electric signals and is within the bandwidth of many such signals. In this study, two different methods were compared for their efficacy in removing 50 Hz noise added to surface electromyogram (EMG) signals free of power line interference. The first was a simple second-order recursive digital notch filter. The second was an approach called spectrum interpolation, in which it is assumed that the magnitude of the original 50 Hz component of the EMG signal can be approximated by interpolation of the amplitude spectrum of the signal. When the spectrum was based on records containing an integer number of cycles of 50 Hz interference, and the frequency resolution was finer than 1 Hz, spectrum interpolation performed similarly to, or significantly better than, the notch filter (p<0.01). It was also possible to make spectrum interpolation more robust than the notch filter. The Pearson squared correlation coefficient r² between clean signals and signals processed using the notch filter was reduced from 0.98 to 0.65 when the interference frequency was increased by 0.5 Hz, but r² for spectrum interpolation at 0.2 Hz resolution was only reduced from 0.99 to 0.85 if spectral values between approximately 49.5 and 50.5 Hz were modified by interpolation.     

ECG-changes in the fetal Iamb during asphyxia in relation to beta-adrenoceptor stimulation and blockade

Progressive changes in the S-T interval of the fetal electrocardiogram (FECG) were studied in 14 lamb fetuses, acutely exteriorized and subjected to graded hypoxia. The aims of the study were to investigate whether beta-adrenoceptor stimulation and hypoxia exerted additive or potentiating effects on the FECG and several cardiovascular parameters and whether the hypoxic changes of the FECG could be blocked by beta-adrenoceptor blocking agents. The FECG changes were studied in order to correlate them with cardiovascular function, as measured by heart rate, mean arterial pressure, end diastolic pressure, maximum dP/dt and combined cardiac output, estimated by the thermodilution method, as well as with blood gases, acid base status, blood lactate and glucose. Injections of small doses (0.02 to 0.4 μg kg^-1 min^-1) of isoprenaline induced the same pattern of changes in the FECG as we have previously recorded during hypoxia. By increasing the isoprenaline dose an increase in the duration of the FECG changes and amplitude of the T-wave changes was obtained. Propranolol was found to completely abolish the FECG changes induced by isoprenaline, as well as by mild hypoxia. During severe hypoxia the FECG changes could not be abolished by propranolol. Our previous findings indicated that the hypoxic changes could be regarded as a sign of myocardial glycolysis. Thus, the present finding that even small doses of isoprenaline given to the fetus, initiates the same pattern of FECG changes corroborate this hypothesis.     

Optimal EMG signal bandwidth and interelectrode distance for the recording of acoustic, electrocutaneous, and photic blink reflexes

The bandwidth for the recording of the orbicularis oculi blink reflex electromyogram (EMG) response is optimal when low-frequency artifacts, such as motion artifacts and cross-talk from other muscles, are maximally suppressed, whereas true EMG signal power is maximally retained. The optimal bandwidth was investigated for acoustic, electrocutaneous, and photic blink reflexes. Reflexes were recorded with varying bandwidth and interelectrode distances of 12 and 36 mm. Power spectra of the EMG signals were calculated and compared with a theoretical spectrum of the uncontaminated EMG signal. For both electrode distances, the optimal bandwidth was on the average 28–500 Hz for acoustic and electrocutaneous blink reflexes and 12–500 Hz for photic blinks. Using photic stimuli, however, a high-pass filter frequency larger than 12 Hz (probably at least 30 Hz) in combination with occlusion of the eye will be necessary to avoid influences of retinal potentials. Given the optimal bandwidth, a larger electrode spacing may be expected to moderately improve the detectability of small blinks in all stimulus conditions.     

The interference threshold of cardiac pacemakers in electric 50 Hz fields

The aim of the study was to determine a 'worst-case' and a 'real-case' interference threshold for implanted cardiac pace-makers (CPM) in electric 50 Hz fields as they appear in high-voltage plants, e.g. beneath high voltage overhead lines. For this purpose the resulting electrical potential distribution within the thorax area of volunteers from an external homogeneous electrical 50 Hz field was measured. Different factors such as different body geometries as well as inspiration and expiration of the lung were considered. Measurements showed that 1 per 1 kV m(-1) unimpaired electrical field strength (RMS) an interference voltage of about 180 microVpp as real-case value an 400 microVpp as worst-case value would occur at the input of a unipolar ventricularly controlled, left pectorally implanted cardiac pacemaker. Therefore, it is possible under worst-case conditions but unlikely under practice-relevant conditions that an implanted cardiac pacemaker is disturbed by present electric 50 Hz fields beneath high voltage overhead lines.     

Multichannel recording of cardiac potentials

An inexpensive system for the simultaneous recording of 256 cardiac electrograms is described. Time division multiplexing of cardiac potentials by digital circuitry provides a single video signal which is recorded on video tape. During playback a demultiplexer reassembles the separate signals. Measured channel characteristics demonstrate a bandwidth of 170 Hz, a noise level of 0·2 mV peak-to-peak, and a crosstalk of −27·5 dB. The design has flexibility in that individual channel bandwidth can be exchanged with the total number of channels according to the dictates of signal requirements. To illustrate this flexibility a second system is described which provides 64 channels whose measured characteristics include an 880 Hz bandwidth, a noise level of 0·14 mV peak-to-peak and a crosstalk of −33 dB.     

High-quality recording of bioelectric events

A multichannel instrumentation amplifier, developed to be used in a miniature universal eight-channel amplifier module, is described. After discussing the specific properties of a bioelectric recording, the difficulties of meeting the demanded specifications with a design based on operational amplifiers are reviewed. Because it proved impossible to achieve the demanded combination of low noise and low power consumption using commercially available operational amplifiers, an amplifier equipped with an input stage with discrete transistors was developed. A new design concept was used to expand the design to a multichannel version with an equivalent input noise voltage of 0·35 μV RMS in a bandwidth of 0·1–100 Hz and a power consumption of 0·6 mW per channel. The results of this study are applied to miniature, universal, eight-channel amplifier modules, manufactured with thick-film production techniques. The modules can be coupled to satisfy the demand for a multiple of eight channels. The low power consumption enables the modules to be used in all kinds of portable and telemetry measurement systems and simplifies the power supply in stationary measurement systems.     

Detection of fetal electrocardiogram signals using matched filters with adaptive normalisation

The authors discuss the application of matched filters to the detection of R-waves in fetal electrocardiogram (FECG) data, recorded during labour using a scalp electrode. When using the basic matched filter, one correlates a template representing the clean signal with the noisy signal. This method is optimal when the underlying noise is white in nature. However, it is known that false detection of R-waves can occur in the presence of extraneous peaks which have a similar shape to the fetal R-wave. It is proposed to switch between two different normalisations of the impulse response of the matched filter to alleviate this problem. When the signal-to-noise ratio is lower than a predetermined threshold, then normalisation to the geometric mean of the template and noisy data energies is carried out, otherwise only normalisation to the template energy is made. In the former case, the background noise and spikes that are larger than the underlying FECG are attenuated, hence increasing the probability of detection of the R-waves. In the latter case, noise which has a lower amplitude than the underlying R-wave, is reduced. The effectiveness of this method is demonstrated by application to scalp electrode data.