Fetal Signals Research Articles

Developmental-status-aware transcriptional decomposition establishes a cell state panorama of human cancers

BackgroundCancer cells evolve under unique functional adaptations that unlock transcriptional programs embedded in adult stem and progenitor-like cells for progression, metastasis, and therapeutic resistance. However, it remains challenging to quantify the stemness-aware cell state of a tumor based on its gene expression profile.MethodsWe develop a developmental-status-aware transcriptional decomposition strategy using single-cell RNA-sequencing-derived tissue-specific fetal and adult cell signatures as anchors. We apply our method to various biological contexts, including developing human organs, adult human tissues, experimentally induced differentiation cultures, and bulk human tumors, to benchmark its performance and to reveal novel biology of entangled developmental signaling in oncogenic processes.ResultsOur strategy successfully captures complex dynamics in developmental tissue bulks, reveals remarkable cellular heterogeneity in adult tissues, and resolves the ambiguity of cell identities in in vitro transformations. Applying it to large patient cohorts of bulk RNA-seq, we identify clinically relevant cell-of-origin patterns and observe that decomposed fetal cell signals significantly increase in tumors versus normal tissues and metastases versus primary tumors. Across cancer types, the inferred fetal-state strength outperforms published stemness indices in predicting patient survival and confers substantially improved predictive power for therapeutic responses.ConclusionsOur study not only provides a general approach to quantifying developmental-status-aware cell states of bulk samples but also constructs an information-rich, biologically interpretable, cell-state panorama of human cancers, enabling diverse translational applications.

Cell type and cell signaling innovations underlying mammalian pregnancy.

The fetal-maternal interface is one of the most intense loci of cell-cell signaling in the human body. Invasion of cells from the fetal placenta into the uterus, and the corresponding transformation of maternal tissues called decidualization, first evolved in the stem lineage of eutherian mammals( 1 , 2 ). Single-cell studies of the human fetal-maternal interface have provided new insight into the cell type diversity and cell-cell interactions governing this chimeric organ( 3-5 ). However, the fetal-maternal interface is also one of the most rapidly evolving, and hence most diverse, characters among mammals( 6 ), and an evolutionary analysis is missing. Here, we present and compare single-cell data from the fetal-maternal interface of species bracketing key events in mammal phylogeny: a marsupial (opossum, Monodelphis domestica ), the afrotherian Tenrec ecaudatus, and four Euarchontoglires - guinea pig and mouse (Rodentia) together with recent macaque and human data (primates) ( 4 , 5 , 7 ). We infer cell type homologies, identify a gene-expression signature of eutherian invasive trophoblast conserved over 99 million years, and discover a predecidual cell in the tenrec which suggests stepwise evolution of the decidual stromal cell. We reconstruct ancestral cell signaling networks, revealing the integration of fetal cell types into the interface. Finally, we test two long-standing theoretical predictions, the disambiguation hypothesis( 8 ) and escalation hypothesis( 9 ), at transcriptome-wide scale, finding divergence between fetal and maternal signaling repertoires but arms race dynamics restricted to a small subset of ligand-receptor pairs. In so doing, we trace the co-evolutionary history of cell types and their signaling across mammalian viviparity.

Nutraceuticals in Pregnancy: A Special Focus on Probiotics.

The placenta is crucial to fetal development and performs vital functions such as nutrient exchange, waste removal and hormone regulation. Abnormal placental development can lead to conditions such as fetal growth restriction, pre-eclampsia and stillbirth, affecting both immediate and long-term fetal health. Placental development is a highly complex process involving interactions between maternal and fetal components, imprinted genes, signaling pathways, mitochondria, fetal sexomes and environmental factors such as diet, supplementation and exercise. Probiotics have been shown to make a significant contribution to prenatal health, placental health and fetal development, with associations with reduced risk of preterm birth and pre-eclampsia, as well as improvements in maternal health through effects on gut microbiota, lipid metabolism, vaginal infections, gestational diabetes, allergic diseases and inflammation. This review summarizes key studies on the influence of dietary supplementation on placental development, with a focus on the role of probiotics in prenatal health and fetal development.

Adiponectin and Leptin during Pregnancy: A Systematic Review of Their Association with Pregnancy Disorders, Fetal Growth and Placental Function

During pregnancy, the adipokines leptin and adiponectin can affect placental nutrient transport and inflammatory pathways, potentially leading to altered fetal growth and pregnancy complications including gestational diabetes mellitus (GDM) and preeclampsia (PE). The aim of this systematic review is to gather and analyze research on maternal circulating leptin and adiponectin levels and their relationship to adverse pregnancy and birth outcomes. Additionally, it seeks to determine whether these hormones are linked to alterations in placental transporters and cell signaling pathways. PubMed and MEDLINE were systematically searched to include studies published between 2012 and 2022. All primary data studies reporting serum adiponectin and/or leptin, placental mRNA and protein levels of related transporters, and adverse birth outcomes were eligible. The current systematic review encompasses a total of 14 articles. Abnormal serum maternal leptin and adiponectin levels were associated with changes in fetal growth and placental cellular signaling and nutrient transporters. A majority of studies associated elevated maternal leptin and reduced adiponectin with fetal overgrowth, although this relationship was not consistent and may be complicated when other pathologies are present. The effects of maternal leptin and adiponectin on fetal growth may be driven by placental adaptation in nutrient transporters and mitochondria. Future studies should determine if the placental effects of leptin and adiponectin that have been found in models have mechanistic roles in human pregnancy.

The Effects of Maternal Nutrient Restriction during Mid to Late Gestation with Realimentation on Fetal Metabolic Profiles in the Liver, Skeletal Muscle, and Blood in Sheep.

Poor maternal nutrition during gestation negatively affects offspring growth and metabolism. To evaluate the impact of maternal nutrient restriction and realimentation on metabolism in the fetal liver, skeletal muscle, and circulation, on day 50 of gestation, ewes (n = 48) pregnant with singletons were fed 100% (CON) or 60% (RES) of requirements until day 90 of gestation, when a subset of ewes (n = 7/treatment) were euthanized, and fetal samples were collected. The remaining ewes were maintained on a current diet (CON-CON, n = 6; RES-RES, n = 7) or switched to an alternative diet (CON-RES, RES-CON; n = 7/treatment). On day 130 of gestation, the remaining ewes were euthanized, and fetal samples were collected. Fetal liver, longissimus dorsi (LD), and blood metabolites were analyzed using LC-MS/MS, and pathway enrichment analysis was conducted using MetaboAnalyst. Then, 600, 518, and 524 metabolites were identified in the liver, LD, and blood, respectively, including 345 metabolites that were present in all three. Nutrient restriction was associated with changes in amino acid, carbohydrate, lipid, and transulfuration/methionine metabolic pathways, some of which were alleviated by realimentation. Fetal age also affected metabolite abundance. The differential abundance of metabolites involved in amino acid, methionine, betaine, and bile acid metabolism could impact fetal epigenetic regulation, protein synthesis, lipid metabolism, and signaling associated with glucose and lipid metabolism.

Sensitivity analysis of transabdominal fetal pulse oximetry using MRI-based simulations.

Transabdominal fetal pulse oximetry offers a promising approach to improve fetal monitoring and reduce unnecessary interventions. Utilizing realistic 3D geometries derived from MRI scans of pregnant women, we conducted photon simulations to determine optimal source-detector configurations for detecting fetal heart rate and oxygenation. Our findings demonstrate the theoretical feasibility of measuring fetal signals at depths up to 30 mm using source-detector (SD) distances greater than 100 mm and wavelengths between 730 and 850 nm. Furthermore, we highlight the importance of customizing SD configurations based on fetal position and maternal anatomy. These insights pave the way for enhanced non-invasive fetal monitoring in clinical application.

Extracting fetal heart signals from Doppler using semi-supervised convolutional neural networks.

Cardiotocography (CTG) measurements are critical for assessing fetal wellbeing during monitoring, and accurate assessment requires well-traceable CTG signals. The current FHR calculation algorithm, based on autocorrelation to Doppler ultrasound (DUS) signals, often results in periods of loss owing to its inability to differentiate signals. We hypothesized that classifying DUS signals by type could be a solution and proposed that an artificial intelligence (AI)-based approach could be used for classification. However, limited studies have incorporated the use of AI for DUS signals because of the limited data availability. Therefore, this study focused on evaluating the effectiveness of semi-supervised learning in enhancing classification accuracy, even in limited datasets, for DUS signals. Data comprising fetal heartbeat, artifacts, and two other categories were created from non-stress tests and labor DUS signals. With labeled and unlabeled data totaling 9,600 and 48,000 data points, respectively, the semi-supervised learning model consistently outperformed the supervised learning model, achieving an average classification accuracy of 80.9%. The preliminary findings indicate that applying semi-supervised learning to the development of AI models using DUS signals can achieve high generalization accuracy and reduce the effort. This approach may enhance the quality of fetal monitoring.

Fetal electrocardiogram signal extraction based on multi-scale residual shrinkage U-Net

In the extraction of fetal electrocardiogram (ECG) signal, due to the unicity of the scale of the U-Net same-level convolution encoder, the size and shape difference of the ECG characteristic wave between mother and fetus are ignored, and the time information of ECG signals is not used in the threshold learning process of the encoder's residual shrinkage module. In this paper, a method of extracting fetal ECG signal based on multi-scale residual shrinkage U-Net model is proposed. First, the Inception and time domain attention were introduced into the residual shrinkage module to enhance the multi-scale feature extraction ability of the same level convolution encoder and the utilization of the time domain information of fetal ECG signal. In order to maintain more local details of ECG waveform, the maximum pooling in U-Net was replaced by Softpool. Finally, the decoder composed of the residual module and up-sampling gradually generated fetal ECG signals. In this paper, clinical ECG signals were used for experiments. The final results showed that compared with other fetal ECG extraction algorithms, the method proposed in this paper could extract clearer fetal ECG signals. The sensitivity, positive predictive value, and F1 scores in the 2013 competition data set reached 93.33%, 99.36%, and 96.09%, respectively, indicating that this method can effectively extract fetal ECG signals and has certain application values for perinatal fetal health monitoring.

Advancing non-invasive fetal health monitoring: A time–frequency approach to extracting fetal electrocardiogram signals

The fetal electrocardiogram (FECG) provides an important way for clinicians to assess fetal health and to monitor pregnancy conditions. Although FECG can be captured through abdominal electrocardiogram (AECG) recordings from pregnant women, it often receives interference from maternal cardiac activity and other external sources. Particularly, as the high-amplitude maternal electrocardiogram (MECG) typically obscures the waveform of the FECG, isolating FECG from the AECG signals is challenging. Nevertheless, the differences in frequency patterns between MECG and FECG suggest a potential way for reconstructing FECG by combining the frequency and amplitude information. To this end, we proposed a novel method for separating FECG signals from the AECG via the time–frequency representation. We designed an encoder–decoder architecture based hierarchical multiscale-aware network to capture features embedded in both the time and frequency domains. Equipped with a holistic attention-aware module (HAA), the model extracts and consolidates fetal heart activity-related features in the time–frequency domain. Further introducingan attention gate module (AG) to emphasize key features during encoder-to-decoder transmission, the model enhances the quality and efficiency of FECG extraction without any redundancy in high-quality feature mapping. The model was trained using a hybrid time-domain and frequency-domain loss function and validated on two public databases. Excellent performances, such as an average Pearson correlation coefficient of 0.85 for estimating the quality of FECG signals, were observed demonstrating its excellent performance and providing a promising non-invasive method for monitoring fetal health. These findings indicate that the proposed method effectively captures the amplitude differences between fetal and maternal heartbeats across various frequency components within the AECG, leading to the extraction of high-quality FECG signals.

Compression of electrocardiogram signals using compressive sensing technique based on curvelet transform toward medical applications

AbstractElectrocardiogram (ECG) signals can be monitored from many patients based on healthcare systems. To enhance these systems, the ECG signals should be collected and then stored in a cloud platform for later analysis. Hence, ECG signals can be utilized to diagnose heart diseases. However, the ECG signals require great internet capacity. So, compression techniques can be implemented to reduce a memory storage capacity for these signals. One of the potential compression techniques is the compressive sensing (CS). This paper proposes a CS technique to compress ECG signals. This technique is used to reduce sampling rates of the ECG signals to be less than the Nyquist rate. Moreover, a framework is suggested for the compression of maternal and fetal ECG signals. The compression of these signals is based on the curvelet transform (CT) to produce sparsity in ECG signals. The MIT-BIH database are utilized for testing the ECG signals. This database includes several ECG signals with various sampling rates, such as aberrant and normal signals. The proposed CS technique achieved a compression ratio (CR) of 15.7 with an accuracy of 98.2%. Also, a percentage root mean difference (PRD) is utilized to calculate the performance of the reconstructed ECG signals. The achieved value of the PRD is 2.0.

The Development and Implementation of Innovative Blind Source Separation Techniques for Real-Time Extraction and Analysis of Fetal and Maternal Electrocardiogram Signals.

This article presents an innovative approach to analyzing and extracting electrocardiogram (ECG) signals from the abdomen and thorax of pregnant women, with the primary goal of isolating fetal ECG (fECG) and maternal ECG (mECG) signals. To resolve the difficulties related to the low amplitude of the fECG, various noise sources during signal acquisition, and the overlapping of R waves, we developed a new method for extracting ECG signals using blind source separation techniques. This method is based on independent component analysis algorithms to detect and accurately extract fECG and mECG signals from abdomen and thorax data. To validate our approach, we carried out experiments using a real and reliable database for the evaluation of fECG extraction algorithms. Moreover, to demonstrate real-time applicability, we implemented our method in an embedded card linked to electronic modules that measure blood oxygen saturation (SpO2) and body temperature, as well as the transmission of data to a web server. This enables us to present all information related to the fetus and its mother in a mobile application to assist doctors in diagnosing the fetus's condition. Our results demonstrate the effectiveness of our approach in isolating fECG and mECG signals under difficult conditions and also calculating different heart rates (fBPM and mBPM), which offers promising prospects for improving fetal monitoring and maternal healthcare during pregnancy.

Transcutaneous Discrimination of Fetal Heart Rate from Maternal Heart Rate: A Fetal Oximetry Proof-of-Concept.

Intrapartum care uses electronic fetal heart rate monitoring (EFHRM) for over 50years to indirectly assess fetal oxygenation. However, this approach has been associated with an increase in cesarean delivery rates and limited improvements in neonatal hypoxic outcome. To address these shortcomings, a novel transabdominal fetal pulse oximeter (TFO) is being developed to provide an objective measurement of fetal oxygenation. Previous studies have evaluated the performance of TFO on pregnant ewe. Building on the animal model, this study aims to determine whether TFO can successfully capture human fetal heart rate (FHR) signals during non-stress testing (NST) as a proof-of-concept. Eight ongoing pregnancies meeting specific inclusion criteria (18-40years old, singleton, and at least 36weeks' gestation) were enrolled with consent. Each study session was 15 to 20min long. Reference maternal heart rate (MHR) and FHR were obtained using finger pulse oximetry and cardiotocography for subsequent comparison. The overall root-mean-square error was 9.7BPM for FHR and 4.4 for MHR, while the overall mean-absolute error was 7.6BPM for FHR and 1.8 for MHR. Bland-Altman analysis displayed a mean bias ± standard deviation between TFO and reference of -3.9 ± 8.9BPM, with limits of agreement ranging from -21.4 to 13.6 BPM. Both maternal and fetal heart rate measurements obtained from TFO exhibited a p-value < 0.001, showing significant correlation with the reference. This proof-of-concept study successfully demonstrates that TFO can accurately differentiate maternal and fetal heart signals in human subjects. This achievement marks the initial step towards enabling fetal oxygen saturation measurement in humans using TFO.

Enhancing Fetal Electrocardiogram Signal Extraction Accuracy through a CycleGAN Utilizing Combined CNN-BiLSTM Architecture.

The fetal electrocardiogram (FECG) records changes in the graph of fetal cardiac action potential during conduction, reflecting the developmental status of the fetus in utero and its physiological cardiac activity. Morphological alterations in the FECG can indicate intrauterine hypoxia, fetal distress, and neonatal asphyxia early on, enhancing maternal and fetal safety through prompt clinical intervention, thereby reducing neonatal morbidity and mortality. To reconstruct FECG signals with clear morphological information, this paper proposes a novel deep learning model, CBLS-CycleGAN. The model's generator combines spatial features extracted by the CNN with temporal features extracted by the BiLSTM network, thus ensuring that the reconstructed signals possess combined features with spatial and temporal dependencies. The model's discriminator utilizes PatchGAN, employing small segments of the signal as discriminative inputs to concentrate the training process on capturing signal details. Evaluating the model using two real FECG signal databases, namely "Abdominal and Direct Fetal ECG Database" and "Fetal Electrocardiograms, Direct and Abdominal with Reference Heartbeat Annotations", resulted in a mean MSE and MAE of 0.019 and 0.006, respectively. It detects the FQRS compound wave with a sensitivity, positive predictive value, and F1 of 99.51%, 99.57%, and 99.54%, respectively. This paper's model effectively preserves the morphological information of FECG signals, capturing not only the FQRS compound wave but also the fetal P-wave, T-wave, P-R interval, and ST segment information, providing clinicians with crucial diagnostic insights and a scientific foundation for developing rational treatment protocols.

Non-Invasive Fetal Well-Being Monitoring Approaches: A Mini Review on Fetal Signal Separation Techniques

Fetal signal separation is vital in producing an accurate interpretation of the health condition of a fetal. In the context of a non-invasive fetal monitoring approach, the signals are acquired from the abdomen of pregnant women. As a result, a mix of maternal and fetal signals is obtained. These maternal and fetal signals are vague, as both signals are interchanged during the signal acquisition stage. Since the signals are overlapped, a signal separation technique must be employed to process the fetal signal for further analysis. This paper presents published studies on applying signal processing techniques involving fetal signal separation. These papers are obtained through a strategy known as the PRISMA technique. The online databases include ACM, Emerald Publishing, IEEE Explore Digital Library, Science Direct, Scopus, and Springer, with published years spanning from 2018 until 2022. Numerous separation techniques were found, such as adaptive filtering, blind source separation (BSS), and alternative approaches. Issues on the existing methods for fetal signal separation are discussed. In addition, the limitations and drawbacks of the research work involving existing fetal signal separation are reviewed in the paper. The potential direction of future research in this field is addressed as well. Based on this mini-review, it can be concluded that noise and ambiguity can still occur in the extracted fetal signals, even when signal processing techniques are applied. In the future, deep learning would be accommodating in improving the efficiency of extracting fetal signals obtained from the non-invasive fetal well-being monitoring technique. Meanwhile, apart from fetal heart rate (fHR) detection, fetal hypoxia can also be another important focus of study for improving fetal well-being monitoring.

A deep learning framework for noninvasive fetal ECG signal extraction.

Introduction: The availability of proactive techniques for health monitoring is essential to reducing fetal mortality and avoiding complications in fetal wellbeing. In harsh circumstances such as pandemics, earthquakes, and low-resource settings, the incompetence of many healthcare systems worldwide in providing essential services, especially for pregnant women, is critical. Being able to continuously monitor the fetus in hospitals and homes in a direct and fast manner is very important in such conditions. Methods: Monitoring the health of the baby can potentially be accomplished through the computation of vital bio-signal measures using a clear fetal electrocardiogram (ECG) signal. The aim of this study is to develop a framework to detect and identify the R-peaks of the fetal ECG directly from a 12 channel abdominal composite signal. Thus, signals were recorded noninvasively from 70 pregnant (healthy and with health conditions) women with no records of fetal abnormalities. The proposed model employs a recurrent neural network architecture to robustly detect the fetal ECG R-peaks. Results: To test the proposed framework, we performed both subject-dependent (5-fold cross-validation) and independent (leave-one-subject-out) tests. The proposed framework achieved average accuracy values of 94.2% and 88.8%, respectively. More specifically, the leave-one-subject-out test accuracy was 86.7% during the challenging period of vernix caseosa layer formation. Furthermore, we computed the fetal heart rate from the detected R-peaks, and the demonstrated results highlight the robustness of the proposed framework. Discussion: This work has the potential to cater to the critical industry of maternal and fetal healthcare as well as advance related applications.

Construction of a comprehensive fetal monitoring database for the study of perinatal hypoxic ischemic encephalopathy

This article describes the methods used to build a large-scale database of more than 250,000 electronic fetal monitoring (EFM) records linked to a comprehensive set of clinical information about the infant, the mother, the pregnancy, labor, and outcome. The database can be used to investigate how birth outcome is related to clinical and EFM features. The main steps involved in building the database were: (1) Acquiring the raw EFM recording and clinical records for each birth. (2) Assigning each birth to an objectively defined outcome class that included normal, acidosis, and hypoxic-ischemic encephalopathy. (3) Removing all personal health information from the EFM recordings and clinical records. (4) Preprocessing the deidentified EFM records to eliminate duplicates, reformat the signals, combine signals from different sensors, and bridge gaps to generate signals in a format that can be readily analyzed. (5) Post-processing the repaired EFM recordings to extract key features of the fetal heart rate, uterine activity, and their relations. (6) Populating a database that links the clinical information, EFM records, and EFM features to support easy querying and retrieval.•A multi-step process is required to build a comprehensive database linking electronic temporal fetal monitoring signals to a comprehensive set of clinical information about the infant, the mother, the pregnancy, labor, and outcome.•The current database documents more than 250,000 births including almost 4,000 acidosis and 400 HIE cases. This represents more than 80% of the births that occurred in 15 Northern California Kaiser Permanente Hospitals between 2011–2019. This is a valuable resource for studying the factors predictive of outcome.•The signal processing code and schemas for the database are freely available. The database will not be permitted to leave Kaiser firewalls, but a process is in place to allow interested investigators to access it.

An Integrated Framework for Assessing the Quality of Non-invasive Fetal Electrocardiography Signals

An Integrated Framework for Assessing the Quality of Non-invasive Fetal Electrocardiography Signals

A novel Discrete Artificial Bee Colony algorithm combined with adaptive filtering to extract Fetal Electrocardiogram signals

The Fetal Electrocardiogram (FECG) signal plays a crucial role in monitoring the health of the fetus, but there are numerous challenges in eliminating the maternal thorax signal and reducing noise interference. This paper proposes a novel objective function that combines a Least Mean Squares (LMS) adaptive filter with a heuristic algorithms to enhance the quality of the extracted FECG signal. To achieve better results, we introduce the Discrete Artificial Bee Colony (DABC) algorithm with a new initialization strategy, a random wavelet basic function strategy, and Gaussian distribution. These improvements enhance global search capabilities and ensure a faster convergence rate. The application of heuristic algorithms can reduce noise signals and provides clearer and more accurate results compared to the traditional LMS filter. Furthermore, the effectiveness of this innovative algorithm is compared with other widely used heuristic algorithms. The experiment results demonstrate that the novel algorithm significantly enhances performance by up to 8% compared to other conventional extraction methods in some indicators.

Optimized systolic array filters with noise classification for extracting FECG

To preserve the lives of both mother and foetus during the early phases of childbirth, heart abnormality diagnosis is essential. In remote places where understanding of maternal care is limited, the death rate caused by carelessness or a failure to detect abnormalities is still a problem. The signal will contain various external or internal noise sources to isolate the key components of the foetal heart rate from the mother's belly during labour. The likelihood that the genuine signal would be misinterpreted and result in a false report will increase in the presence of such noise sources. Although there are many software programs available for extracting the QRS features from foetal ECG signals, it is unavoidable that specialized hardware is required for a significant reduction in both area and power. This paper's main goal is to extract the QRS complex using LDA, then improve Social Spider classifier performance using the suggested TAODV as a distance metric calculator, and then compare against existing methods to discover sounds that are distorting the normal heart rate. Systolic array filter with suggested Glitch Avoidance Circuit employing MUX is simulated using Cadence Virtuoso in 65nm technology to remove noise from the observed QRS complex. Over 100 records with the necessary examples from MIT-BIH Arrhythmia were used in the simulations, and it was discovered that MATLAB 2010b was used to adopt a unique technique for classifying noise. The suggested TAODV-based SSA classifier's accuracy is 96.8%, whereas the accuracy of a filter with a glitch avoidance circuit is 96.13%. The primary benefit of these strategies comprises cutting-edge hardware and computational solutions.

Multichannel high noise level ECG denoising based on adversarial deep learning

This paper proposes a denoising method based on an adversarial deep learning approach for the post-processing of multi-channel fetal electrocardiogram (ECG) signals. As it’s well known, noise leads to misinterpretations of fetal ECG signals and thus limits the use of fetal electrocardiography for healthcare applications. Therefore, denoising algorithms are essential for the exploitation of non-invasive fetal ECG. The proposed method is based on the combination of three end-to-end trained sub-networks to convert noisy fetal ECG signals into clean signals. The first two sub-networks are linked by skip connections and form a deep convolutional network that downsamples the noisy signals into a latent representation and subsequently upsamples this latent representation to recover clean signals. The third sub-network aims to boost the decoder sub-network to generate realistic clean signals. Experiments carried out on synthetic and real data showed that the proposed method improved by the signal-to-noise (SNR) of fetal ECG signals with input SNR ranging from -,30 to 0 dB by an average of 20 dB, and improve fetal signal quality by significantly increasing the number of true detected QRS complexes and halving QRS complex detection errors.