Automatic signal quality assessment of raw trans-abdominal biopotential recordings for non-invasive fetal electrocardiography.

Introduction: Wearable monitoring systems for non-invasive multi-channel fetal electrocardiography (fECG) can support fetal surveillance and diagnosis during pregnancy, thus enabling prompt treatment. In these embedded systems, power saving is the key to long-term monitoring. In this regard, the computational burden of signal processing methods implemented for the fECG extraction from the multi-channel trans-abdominal recordings plays a non-negligible role. In this work, a supervised machine-learning approach for the automatic selection of the most informative raw abdominal recordings in terms of fECG content, i.e., those potentially leading to good-quality, non-invasive fECG signals from a low number of channels, is presented and evaluated. Methods: For this purpose, several signal quality indexes from the scientific literature were adopted as features to train an ensemble tree classifier, which was asked to perform a binary classification between informative and non-informative abdominal channels. To reduce the dimensionality of the classification problem, and to improve the performance, a feature selection approach was also implemented for the identification of a subset of optimal features. 10336 5-s long signal segments derived from a real dataset of multi-channel trans-abdominal recordings acquired from 55 voluntary pregnant women between the 21st and the 27th week of gestation, with healthy fetuses, were adopted to train and test the classification approach in a stratified 10-time 10-fold cross-validation scheme. Abdominal recordings were firstly pre-processed and then labeled as informative or non-informative, according to the signal-to-noise ratio exhibited by the extracted fECG, thus producing a balanced dataset of bad and good quality abdominal channels. Results and Discussion: Classification performance revealed an accuracy above 86%, and more than 88% of those channels labeled as informative were correctly identified. Furthermore, by applying the proposed method to 50 annotated 24-channel recordings from the NInFEA dataset, a significant improvement was observed in fetal QRS detection when only the channels selected by the proposed approach were considered, compared with the use of all the available channels. As such, our findings support the hypothesis that performing a channel selection by looking directly at the raw abdominal signals, regardless of the fetal presentation, can produce a reliable measurement of fetal heart rate with a lower computational burden.

Impact of pulsed-wave-Doppler velocity-envelope tracing techniques on classification of complete fetal cardiac cycles.

Fetal echocardiography is an operator-dependent examination technique requiring a high level of expertise. Pulsed-wave Doppler (PWD) is often used as a reference for the mechanical activity of the heart, from which several quantitative parameters can be extracted. These aspects suggest the development of software tools that can reliably identify complete and clinically meaningful fetal cardiac cycles that can enable their automatic measurement. Several scientific works have addressed the tracing of the PWD velocity envelope. In this work, we assess the different steps involved in the signal processing chains that enable PWD envelope tracing. We apply a supervised classifier trained on envelopes traced by different signal processing chains for distinguishing complete and measurable PWD heartbeats from incomplete or malformed ones, which makes it possible to determine the impact of each of the different processing steps on the detection accuracy. In this study, we collected 43 images and labeled 174,319 PWD segments from 25 pregnant women volunteers. By considering seven envelope tracing techniques and the 23 different processing steps involved in their implementation, the results of our study reveal that, compared to the steps investigated in most other works, those that achieve binarisation and envelope extraction are significantly more important (p < 0.05). The best approaches among those studied enabled greater than 98% accuracy on our large manually annotated dataset.

A non-invasive multimodal foetal ECG-Doppler dataset for antenatal cardiology research.

Non-invasive foetal electrocardiography (fECG) continues to be an open topic for research. The development of standard algorithms for the extraction of the fECG from the maternal electrophysiological interference is limited by the lack of publicly available reference datasets that could be used to benchmark different algorithms while providing a ground truth for foetal heart activity when an invasive scalp lead is unavailable. In this work, we present the Non-Invasive Multimodal Foetal ECG-Doppler Dataset for Antenatal Cardiology Research (NInFEA), the first open-access multimodal early-pregnancy dataset in the field that features simultaneous non-invasive electrophysiological recordings and foetal pulsed-wave Doppler (PWD). The dataset is mainly conceived for researchers working on fECG signal processing algorithms. The dataset includes 60 entries from 39 pregnant women, between the 21st and 27th week of gestation. Each dataset entry comprises 27 electrophysiological channels (2048 Hz, 22 bits), a maternal respiration signal, synchronised foetal trans-abdominal PWD and clinical annotations provided by expert clinicians during signal acquisition. MATLAB snippets for data processing are also provided.

Annotated real and synthetic datasets for non-invasive foetal electrocardiography post-processing benchmarking.

Non-invasive foetal electrocardiography (fECG) can be obtained at different gestational ages by means of surface electrodes applied on the maternal abdomen. The signal-to-noise ratio (SNR) of the fECG is usually low, due to the small size of the foetal heart, the foetal-maternal compartment, the maternal physiological interferences and the instrumental noise. Even after powerful fECG extraction algorithms, a post-processing step could be required to improve the SNR of the fECG signal. In order to support the researchers in the field, this work presents an annotated dataset of real and synthetic signals, which was used for the study "Wavelet Denoising as a Post-Processing Enhancement Method for Non-Invasive Foetal Electrocardiography" [1]. Specifically, 21 15 s-long fECG, dual-channel signals obtained by multi-reference adaptive filtering from real electrophysiological recordings were included. The annotation of the foetal R peaks by an expert cardiologist was also provided. Recordings were performed on 17 voluntary pregnant women between the 21st and the 27th week of gestation. The raw recordings were also included for the researchers interested in applying a different fECG extraction algorithm. Moreover, 40 10 s-long synthetic non-invasive fECG were provided, simulating the electrode placement of one of the abdominal leads used for the real dataset. The annotation of the foetal R peaks was also provided, as generated by the FECGSYN tool used for the signals' creation. Clean fECG signals were also included for the computation of indexes of signal morphology preservation. All the signals are sampled at 2048 Hz. The data provided in this work can be used as a benchmark for fECG post-processing techniques but can also be used as raw signals for researchers interested in foetal QRS detection algorithms and fECG extraction methods.

A Novel Tool for Non-Invasive Fetal Electrocardiography Research: the NInFEA Dataset.

In this work, a novel open-source dataset for noninvasive fetal electrocardiography research is presented. It is composed of 60 high-quality electrophysiological recordings acquired between the 21st and the 27th weeks of gestation. For each acquisition, whose average duration is 30.5 s, 24 unipolar abdominal leads and three bipolar thoracic leads were included, along with a maternal respiration signal collected by a thoracic resistive belt. The chosen electrodes positioning map allows reproducing up to ten setups presented in the scientific literature. Each biopotential recording was acquired synchronously with the corresponding fetal cardiac pulsed-wave Doppler (PWD) signal, to provide complete information about the fetal cardiac cycle, both from the electrical and mechanical point of view.This is the first dataset allowing the non-invasive fetal ECG analysis even in early pregnancies with a ground truth about the fetal heart activity, given by the PWD signal. For this reason, it can be used to assess fetal ECG extraction algorithms requiring multiple channels, eventually including maternal references. This dataset is being released on Physionet by the end of June 2020 and will be continuously improved in the framework of the Non-Invasive Fetal ECG Analysis (NInFEA) project of the University of Cagliari (Italy).

Wavelet denoising as a post-processing enhancement method for non-invasive foetal electrocardiography.

BACKGROUND AND OBJECTIVE: The detection of a clean and undistorted foetal electrocardiogram (fECG) from non-invasive abdominal recordings is an open research issue. Several physiological and instrumental noise sources hamper this process, even after that powerful fECG extraction algorithms have been used. Wavelet denoising is widely used for the improvement of the SNR in biomedical signal processing. This work aims to systematically assess conventional and unconventional wavelet denoising approaches for the post-processing of fECG signals by providing evidence of their effectiveness in improving fECG SNR while preserving the morphology of the signal of interest. METHODS: The stationary wavelet transform (SWT) and the stationary wavelet packet transform (SWPT) were considered, due to their different granularity in the sub-band decomposition of the signal. Three thresholds from the literature, either conventional (Minimax and Universal) and unconventional, were selected. To this aim, the unconventional one was adapted for the first time to SWPT by trying different approaches. The decomposition depth was studied in relation to the characteristics of the fECG signal. Synthetic and real datasets, publicly available for benchmarking and research, were used for quantitative analysis in terms of noise reduction, foetal QRS detection performance and preservation of fECG morphology. RESULTS: The adoption of wavelet denoising approaches generally improved the SNR. Interestingly, the SWT methods outperformed the SWPT ones in morphology preservation (p<0.04) and SNR (p<0.0003), despite their coarser granularity in the sub-band analysis. Remarkably, the Han et al. threshold, adopted for the first time for fECG processing, provided the best quality improvement (p<0.003). CONCLUSIONS: The findings of our systematic analysis suggest that particular care must be taken when selecting and using wavelet denoising for non-invasive fECG signal post-processing. In particular, despite the general noise reduction capability, signal morphology can be significantly altered on the basis of the parameterization of the wavelet methods. Remarkably, the adoption of a finer sub-band decomposition provided by the wavelet packet was not able to improve the quality of the processing.

Automatic detection of complete and measurable cardiac cycles in antenatal pulsed-wave Doppler signals.

BACKGROUND AND OBJECTIVE: Pulsed-wave Doppler (PWD) echocardiography is the primary tool for antenatal cardiological diagnosis. Based on it, different measurements and validated reference parameters can be extracted. The automatic detection of complete and measurable cardiac cycles would represent a useful tool for the quality assessment of the PWD trace and the automated analysis of long traces. METHODS: This work proposes and compares three different algorithms for this purpose, based on the preliminary extraction of the PWD velocity spectrum envelopes: template matching, supervised classification over a reduced set of relevant waveshape features, and supervised classification over the whole waveshape potentially representing a cardiac cycle. A custom dataset comprising 43 fetal cardiac PWD traces (174,319 signal segments) acquired on an apical five-chamber window was developed and used for the assessment of the different algorithms. RESULTS: The adoption of a supervised classifier trained with the samples representing the upper and lower envelopes of the PWD, with additional features extracted from the image, achieved significantly better results (p < 0.0001) than the other algorithms, with an average accuracy of 98% ± 1% when using an SVM classifier and a leave-one-subject-out cross-validation. Further, the robustness of the results with respect to the classifier model was proved. CONCLUSIONS: The results reveal excellent detection performance, suggesting that the proposed approach can be adopted for the automatic analysis of long PWD traces or embedded in ultrasound machines as a first step for the extraction of measurements and reference clinical parameters.

[Hypnosis in transesophageal echocardiography. The experience in a Pediatric Cardiology and Congenital Heart Disease Unit]. / L'ipnosi nell'ecocardiografia transesofagea. L'esperienza di un Centro di Cardiologia Pediatrica e delle Cardiopatie Congenite.

BACKGROUND: The purpose of this study was to use hypnosis in patients with congenital heart disease undergoing transesophageal echocardiography (TEE). METHODS: From January 2016 to July 2017, 50 adult patients undergoing TEE were randomly assigned to two groups: TEE in hypnosis (n = 23), TEE in sedation (n = 27). Vital parameters (heart rate [HR], blood pressure [BP], oxygen saturation [SO2] before, during and after the procedure) and drug administration were recorded. The State-Trait Anxiety Inventory was performed before and after TEE, the memory and experience of TEE through a structured interview were assessed. RESULTS: All patients in the hypnosis group performed TEE without any sedation. As for anxiety before TEE, no significant differences were observed between groups; after TEE all patients were less anxious than at the beginning (p<0.001) with a greater decrease in patients of the hypnosis group (p<0.001). Before TEE, there were no significant differences also in HR, BP and SO2. During TEE in both groups a similar increase in HR and BP was found (p<0.001), whereas SO2 values remained stable. In the responses to the structured interview, 94% of patients in the sedation group remembered everything vs 36% of the hypnosis group (p<0.05). No differences were found in the other answers between the two groups. CONCLUSIONS: Hypnosis in TEE is useful to improve the emotional experience of patients with congenital heart disease.

Systematic analysis of single- and multi-reference adaptive filters for non-invasive fetal electrocardiography.

Non-invasive fetal electrocardiography (ECG) has been a research challenge for the past few decades. Due to instrumental noise and the spectral overlap of the maternal ECG signal, the signal-to-noise ratio for fetal ECG is very low. Various techniques have been proposed for cancelling the maternal ECG signal and extracting the fetal QRS complex from non-invasive abdominal recordings. Of these, adaptive filters enable satisfactory extraction when there is only a limited number of signal channels available, but the extraction quality is strongly dependent on the electrode placement. In this work, we systematically analyze this issue by comparing single- and multi-reference implementations of QRD-recursive least square (RLS) adaptive filters and evaluating their performances on real and simulated data in terms of the signal-to-interference ratio (SIR), maternal ECG attenuation, and fetal-QRS-complex detection accuracy. Beyond demonstrating the expected superior performance of the multi-reference version (p < 0.05) with respect to all metrics, except the QRS detection accuracy on synthetic data, we also analyze in detail the effectiveness of this technique with different lead orientations with respect to the correct interpretation of the adopted quality indexes. The results reveal that the single-reference approach, which is preferred when only the fetal heart rate is of interest, cannot produce a signal that has acceptable fetal QRS detection accuracy, regardless of the reference lead selection.

Comparison of Single- and Multi-reference QRD-RLS adaptive filter for non-invasive fetal electrocardiography.

Non-invasive fetal electrocardiography (ECG) would allow accessing very relevant information on fetal cardiac function, especially for arrhythmias. However, the signal-to-noise ratio is significantly low, since fetal ECG is embedded in instrumental noise and spectrally overlapping maternal electrophysiological interferences. Among the different techniques proposed in the scientific literature, some variants of adaptive filters have been proposed for maternal ECG cancellation and fetal QRS complex enhancement. Such techniques encompass approaches using one or more reference signals, which is an important aspect for the development of accurate and unobtrusive monitoring systems.In this work, this aspect is systematically analyzed by comparing single- and multi-reference implementations of the QRD-RLS adaptive filter, and by challenging them in the fetal ECG enhancement on three abdominal leads differently oriented in space. The performance is assessed on real data in terms of signal-to-interference ratio, detection of fetal QRS complexes and maternal ECG attenuation. Multi-reference implementation reveals its superiority, whereas the single-reference implementation suffers from the electrodes positioning and cannot be trustily used even for the fetal heart rate only on the adopted dataset.

Automatic Recognition of Complete Atrioventricular Activity in Fetal Pulsed-Wave Doppler Signals.

Echocardiography is the gold standard for antenatal cardiological assessment. However, the adoption of this technique is challenging, since it is intrinsically operator-dependent and because of the different confounding factors related to the fetal heart size, the fetal movements and the ultrasound artifacts. Among the different options, fetal echocardiography is widely used, concurring to an early diagnosis of several cardiac pathologies. In this work, a neural network-based algorithm targeted at the identification of the most important features of Doppler fetal echocardiography videos is presented and evaluated on real signals. Compared to other approaches, the proposed algorithm works on a couple of ID signals, representing the pulse-wave Doppler envelope extracted from the video, thus preserving a Iightweight approach. For the validation, a small dataset was created, including recordings from five voluntary pregnant women 21st to 27th gestational week), for a total of 20 records, 10 seconds each. The dataset was annotated by an expert cardiologist in order to identify the epochs of the signal where a complete readable cardiac cycle could be identified. The performance of the method was evaluated through a 5-fold cross-validation. An average accuracy up to 88% was obtained, confirming the validity of the proposed approach and paving the way to future improvements of the technique.