Phase‐rectified signal averaging: correlation between two monitors and relationship with short‐term variation of fetal heart rate

Abstract

ABSTRACTObjectivesTo establish the correlation between phase‐rectified signal averaging (PRSA) outputs obtained from a novel self‐applicable non‐invasive fetal electrocardiography (NIFECG) monitor with those from computerized cardiotocography (cCTG). A secondary objective was to evaluate the potential for remote assessment of fetal wellbeing by determining the relationship between PRSA and short‐term variation (STV).MethodsThis was a prospective observational study of women with a singleton pregnancy over 28 + 0 weeks' gestation attending a London teaching hospital for cCTG assessment. Participants underwent concurrent cCTG and NIFECG monitoring for up to 60 min. Averaged accelerative (AAC) and decelerative (ADC) capacities and STV were derived by postprocessing and filtration of signals, generating fully (F) and partially (P) filtered results. Linear correlation and accuracy and precision analysis were performed to assess the relationship between PRSA outputs from cCTG and NIFECG, using varying anchor thresholds, and their association with STV.ResultsA total of 306 concurrent cCTG and NIFECG traces were collected from 285 women. F‐filtered NIFECG PRSA (eAAC/eADC) results were generated from 65% of traces, whereas cCTG PRSA (cAAC/cADC) outputs were generated from all. Strong correlations were observed between cAAC and F‐filtered eAAC (r = 0.879, P < 0.001) and between cADC and F‐filtered eADC (r = 0.895, P < 0.001). NIFECG anchor detection decreased significantly with increasing signal loss, and NIFECG PRSA indices showed considerable deviation from those of cCTG when derived from traces in which fewer than 100 anchors were detected. Removing anchor filters from NIFECG traces to generate P‐filtered PRSA outputs weakened the correlation (AAC: r = 0.505, P < 0.001; ADC: r = 0.560, P < 0.001). Lowering the anchor threshold to 100 increased the yield of eAAC and eADC outputs to approximately 74%, whilst maintaining strong correlation with cAAC (r = 0.839, P < 0.001) and cADC (r = 0.815, P < 0.001), respectively. Both cAAC and cADC showed a very strong linear relationship with cCTG STV (r = 0.928, P < 0.001 and r = 0.911, P < 0.001, respectively). Similar findings were observed with eAAC (r = 0.825, P < 0.001) and eADC (r = 0.827, P < 0.001).ConclusionsPRSA appears to be a method of fetal assessment equivalent to STV, but, due to its innate ability to eliminate artifacts, it generates interpretable NIFECG traces with high accuracy at a higher rate. These findings raise the possibility of self‐applied at‐home or remote fetal heart‐rate monitoring with automated reporting, thus enabling increased surveillance in high‐risk women without impacting on service demand. © 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.