Maternal Health: Fetal Monitoring

The aim of this PhD work is to address a strong clinical need with a challenging engineering/signal processing problem to monitor babies during delivery.

The project will address the significant weaknesses of the current fetal monitoring system recently addressed in a recent Cochrane review. The project will develop an ECG-based system that has significantly better specificity to identify and monitor novel features of the fetal ECG signal.

 

Background: Four million babies are delivered each year in US hospitals alone [1]. Of those four million deliveries, over 85% are monitored via electronic fetal monitoring [2]. Intrapartum continuous electronic fetal heart rate (FHR) monitoring was introduced to  identify and respond to intrapartum fetal hypoxia/acidosis promptly to avoid fetal brain injury. The technology monitors the heart rate of both the baby and mother, and analysed in combination with uterine contractions, it aims to measure fetal wellbeing. While a number of randomised controlled trials demonstrated a clear benefit of the technology for the reduction of neonatal seizures, the same trials showed very little benefit of the technology in terms of overall perinatal mortality and cerebral palsy [3]. In fact, electronic fetal monitoring was associated with a 63% increase in cesarean delivery and a 15% increase in instrumental vaginal deliveries [5]. These unnecessary obstetric interventions pose significant additional risks for the mother and the newborn, and subsequent costs to the healthcare system.

 

The current state-of-the-art in fetal monitoring is the the STAN system (ST Analysis, Neoventa Sweden). STAN is a system for fetal surveillance that displays the fetal heart rate and information resulting from the computerised analysis of ST interval of the fetal ECG. There are several significant problems with the STAN system, identified during a recent Cochrane review of the clinical benefit of the device [6]: STAN can only be used when the amniotic sac has been broken, and requires an electrode to be placed directly on the baby's scalp;  detects hypoxia using a relatively simple signal processing algorithm, largely based on “ST” waveforms deviating from their normal profile. The STAN algorithm therefore requires good estimates of a normal ECG waveform to use as a comparison. These normal waveforms can be difficult to establish in clinical practice; Most significantly, the poor specificity of the STAN system meant that there was ultimately no difference in the number of caesarean deliveries and little to suggest that babies were in better condition at birth when the STAN system was used, compared to the normal standard of care. This provides clear and certain motivation for the development of a novel approach.

This PhD seeks to address the significant weaknesses of the STAN system by developing an ECG-based system that:

1.  Does not require a large amount of normal reference data to be gathered in advance;

2.  Has significantly better specificity than the STAN algorithm by identifying and monitoring novel features of the fetal ECG signal that are more predictive of fetal acidosis than the simple features used in the STAN system.  

The approach is based on a novel feature of ECG linked to fetal hypoxia identified through a collaboration between NUI Galway and the Mayo Clinic.

Subscribe to our newsletter