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

Abstract

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.

Fetal QRS Detection by Multiple Channel Temporal Pattern Search Applied to the NInFEA Database

“…With regard to other work on the NInFEA dataset, Sulas et al[3] reported a correlation coefficient of 0.89 between their QRS detection algorithm and ultrasound for instantaneous fetal heart rate over all records. In the present work, for the 57 records successfully processed (excluding the last few seconds of a record that did not fall into a 5 s segment), the correlation coefficient for instantaneous heart rate/RR interval was 0.99.…”

“…In the present work, for the 57 records successfully processed (excluding the last few seconds of a record that did not fall into a 5 s segment), the correlation coefficient for instantaneous heart rate/RR interval was 0.99. Sulas et al[3] derived the ultrasound RR intervals through visual inspection whereas, in the present work, that process was performed by algorithm. Keenan et al[11] applied a prior fetal ECG extraction approach [12] to NInFEA to assess reliability of electrode/lead selection; results were reported in terms of a reliability measure based on a +/-10 beats/minute difference between ECG detected peaks.…”

“…The NInFEA dataset includes 60 records, between 7.5 s and 120 s long, corresponding to 39 pregnant women, between the 21st and 27st week of gestation [2,3,4]. Many different fetal positions are represented.…”

Discovery of Single Independent Latent Variable

“…Dataset. We consider the dataset from [54], which is publicly available 2 on PhysioNet [16]. This dataset was recently published and is part of an ongoing effort to establish a benchmark for noninvasive fECG extraction methods.…”

“…The obtained video was converted into a 1D time-series using a processing scheme based on envelope detection. The code for this processing scheme was provided as a Matlab-script by the authors of [54]. For convenience, we uploaded the obtained 1D time-series after applying the provided Matlab-script, and it is available at https: //github.com/ica-demos/ECG/tree/master/Data.…”