Making labor safer for

               women and their babies

Evaluation of the Fetal QT Interval Using Non-Invasive
Fetal ECG Technology

Joachim Behar 1 , Tingting Zhu 1 , Julien Oster 1 , Alisa Niksch 1 , Douglas Mah 3 , Terrence Chun 4 , James Greenberg 5  , Cassandre Tanner 1 , Jessica Harrop 5 , Alexander Van Esbroeck 6 , Amy Alexander 7 , Michele McCarroll 7 , Timothy Drake 7 , Angela Silber 7 , Reza Sameni 8 , Jay Ward 9 , Adam Wolfberg 9 , Gari D. Clifford 1,9

1University of Oxford, United Kingdom, 2Tufts Medical Center, 3Children's Hospital, Boston 4Seattle Children's Hospital, 5Brigham and Women's Hospital, 6University of Michigan, 7Summa Health System,  8Shiraz University,
9Mindchild Medical, North Andover, MA

Objective:

QT-interval abnormalities in newborns are associated with hypoxia, maternal SSRI use, heart block, and Sudden Infant Death Syndrome. We sought to validate measurement of the fetal QT interval in term laboring women using non-invasive fetal electrocardiogram monitor (Mindchild Medical, N.Andover, MA).

 

 

 

 

Methods:

Fetal ECG data were recorded from 22 laboring women at term using the NI-FECG Meridian Monitor (Mindchild, North Andover) and an invasive fetal scalp electrode (FSE) simultaneously

-105 One-minute epochs were selected for analysis

- Three pediatric electrophysiologists
independently annotated individual waveforms and averaged waveforms from each epoch.

Methods (cont'd):

 

 

 

 

 

 

 

 

 

 

 

- To avoid contextual bias, beats were presented to each annotator in a random order ensuring that each beat in the annotation work flow was drawn from a different patient

- The annotation interface was built upon the Physionet Lightwave interface

 

 

Results:

- The median fetal QTc interval measured in this study was 393ms (IQ range 374ms-403ms) which is similar to the QTc interval seen in newborns (400ms +/- 20ms at the fourth day of life1)

- QT intervals measured on the average data taken from the NI-fECG and the FSE showed a close correspondence (see Figure 3). The root mean square error (RMSE) between all corresponding averaged
NI-FECG and FSE beats was 12.8ms
, which compares to an RMSE of 16.1ms observed in equivalent adult QT studies2

1Schwartz et al.  New England J. of Med. 338(24), 1998: 1709-1714

2Moody, George B. et al.  Comp. in Card., 2006:313-316, IEEE

Conclusion:

- Non-invasive fetal ECG technology accurately measures the fetal QT interval on the population studied

- Further research is needed to determine whether this technology can identify women at risk for newborn SSRI withdrawal syndrome or intrauterine fetal demise

- Further research is needed to automate measurement of the QT interval using signal processing algorithms

 

 

Fig 3.

Comparison of annotations performed on an average FSE and NI-fECG monitor ECG beat by three experts. Note the close correspondence between experts on both the FSE and NI-fECG signal.

Fig 1.

Average scalp ECG cycle with QT interval

Fig 2.  One annotated QT interval (shared blue area) on a scalp FECG recording using our bespoke interface based on Physionet's Lightwave tool. The annotation interface can be accessed at: ibme-web5.eng.ox.ac.uk/lightwave/

 

© 2017—All rights reserved.