Coping with limitations of fetal monitoring instrumentation to improve heart rhythm variability assessment

The most commonly used method of fetal monitoring is based on analysis of the fetal heart activity. Computer-aided fetal monitoring enables extraction of information hidden for visual interpretation - the instantaneous fetal heart rate (FHR) variability. The most natural method of obtaining FHR signal is fetal electrocardiography (FECG), where the FHR has a natural form of unevenly spaced time series of events - heart beats detected in FECG. However, because of problems with FECG recording, the today's instrumentation is based on monitoring of mechanical activity of the fetal heart by means of Doppler ultrasound technique. The ultrasound signal periodicity is determined with autocorrelation methods, so the FHR output signal has a form of evenly spaced instantaneous measurements, some of which are incorrect or duplicate. The aim of the work was to develop a correction algorithm for recognition and removal of these invalid values, to reproduce the FHR signal as time series of events. The new algorithm was compared to other known methods basing on the collected research material and defined performance measures. Thanks to the reference FECG signal registered simultaneously, a detailed analysis of algorithms performance at the level of true heart cycles was possible. Additionally, the influence of signal correction on indices describing the instantaneous FHR variability was evaluated. The obtained results showed that although changing the FHR signal form into time series of events improved the accuracy of indices, but in relation to beat-to-beat variability, that accuracy still does not ensure reliable analysis of instantaneous FHR variability. (c) 2020 Published by Elsevier B.V. on behalf of Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences.