Non-invasive fetal ECG analysis

Abstract

Despite the important advances achieved in the field of adult electrocardiography signal processing, the analysis of the non-invasive fetal electrocardiogram (NI-FECG) remains a challenge. Currently no gold standard database exists which provides labelled FECG QRS complexes (and other morphological parameters), and publications rely either on proprietary databases or a very limited set of data recorded from few (or more often, just one) individuals.The PhysioNet/Computing in Cardiology Challenge 2013 enables to tackle some of these limitations by releasing a set of NI-FECG data publicly to the scientific community in order to evaluate signal processing techniques for NI-FECG extraction. The Challenge aim was to encourage development of accurate algorithms for locating QRS complexes and estimating the QT interval in non-invasive FECG signals. Using carefully reviewed reference QRS annotations and QT intervals as a gold standard, based on simultaneous direct FECG when possible, the Challenge was designed to measure and compare the performance of participants' algorithms objectively. Multiple challenge events were designed to test basic FHR estimation accuracy, as well as accuracy in measurement of inter-beat (RR) and QT intervals needed as a basis for derivation of other FECG features.This editorial reviews the background issues, the design of the Challenge, the key achievements, and the follow-up research generated as a result of the Challenge, published in the concurrent special issue of Physiological Measurement.

Figure 1

Prenatal development time-line with key landmark with respect to fetal monitoring. At 20 weeks the heart can be heard without amplification (Sameni and Clifford, 2010), at 18 the heart can be monitored by ultrasound (Peters et al., 2001), the Non invasive FECG (NI-FECG) and FMCG can be recorded from 20 weeks onward (Peters et al., 2001) but the vernix caseosa forms around 28th–32nd weeks and dissolves in 37th–38th weeks in normal pregnancies (Stinstra, 2001) limiting NI-FECG effectiveness recording during this period.

Figure 2

Frequency and temporal overlap of the MECG and FECG signals. (a) From top to bottom: example of maternal chest ECG, fetal scalp ECG and abdominal ECG (AECG). Note that the AECG contains a mixture of both MECG and FECG and that some FQRS are overlapping with the MQRS - temporal overlap. To produce (a) a notch filter at 60 Hz was used to make the FECG visible on the abdominal channel. (b) Power spectral density distribution (Burg method, order 20) for 5 min of scalp electrode ECG and 5 min of adult ECG. Notice the frequency overlap between the adult and fetal ECG signals particularly in the frequency band of the QRS.

Figure 3

Scatter plot of the scores for the Challenge (best scores are in the lower left corner). Scores for set C and B are marked in blue and red, respectively. The score for the sample entry is highlighted in green.