A wearable light-based sensor is being tested during labour to detect when a baby is not getting enough oxygen, aiming to prevent brain injury. Current monitoring relies on cardiotocography (CTG), which tracks the fetal heart rate but is notoriously unreliable—it has a high false positive rate for detecting oxygen deprivation, has increased rates of emergency Caesarean sections, and has not reduced cerebral palsy. Inadequate fetal monitoring accounts for 70% of avoidable brain injuries at birth in the UK. The new sensor, called FetalSense, uses near-infrared light at six wavelengths to measure oxygen levels and mitochondrial function directly in the placenta, rather than inferring them from heart rate patterns. This feasibility study (MERIT) will test whether the device can collect clean data during labour alongside standard monitoring, and will explore how to preprocess that data for machine-learning models that could eventually flag placental compromise in real time. If successful, the sensor could give clinicians a direct, objective measure of placental function during labour, replacing the subjective interpretation of heart rate traces. This would allow faster, more accurate decisions about when to intervene, potentially reducing the number of babies who suffer lifelong disability from oxygen deprivation at birth.
View original technical description
Hypoxic brain injury in newborn infants is a significant global burden1. It occurs due to inadequate fetoplacental oxygenation, usually acutely in labour. Reducing neonatal harm in childbirth is a UK priority, led by the DHSC-funded "Avoiding Brain Injury in Childbirth (ABC)" programme2, in collaboration with the RCOG3, and THIS Institute4. The most recent NHS Resolution report indicates inadequate fetal monitoring is responsible for 70% of cases of avoidable brain injuries at birth5. Cardiotocography (CTG) is the current mainstay of intrapartum fetal monitoring. It relies on operator's interpretation of the rate, variability and acute changes in the fetal heart rate (FHR) as an indicator of developing fetal hypoxia. The clinical aim is to detect fetal compromise promptly in labour and to institute appropriate delivery to prevent acquired brain injury (ABI). However, CTG has a low specificity in detecting fetal hypoxia with a high false positive rate in predicting CP (cerebral palsy), along with poor inter- and intra-observer agreement6. CTG has increased the rate of operative birth without reducing CP7. There is an urgent need for a scalable, easy-to-use and reliable non-invasive monitoring tool during labour that immediately detects changes in fetoplacental function allowing timely clinical decision-making to reduce neonatal ABI8. We have developed a unique wearable Near-infrared Spectroscopy (NIRS) sensor (FetalSense) for continuous monitoring of placental oxygenation and metabolism. The FetalSense unit is a multiwavelength (6-NIR wavelengths), multi-distance wearable non-invasive sensor, first-of-its-kind to quantify placenta function. Commercially available cerebral NIRS oximeters are now part of clinical monitoring in neonatal, paediatric and adult neurocritical care units. FetalSense utilises innovative algorithms (Spatially Resolved Spectroscopy, Dual Slope) in combination with light diffusion optical simulations of the abdomen, uterus and placental structures to achieve in-depth measurements of absolute placental tissue oxygenation and mitochondrial function via quantification of the redox changes in mitochondrial cytochrome-c-oxidase (oxCCO). FetalSense has had public and patient involvement (PPI) input that has shaped both the project idea and design iterations (see Figure 1 for progression from the initial design (1A) to the current version (1E)). Proof-of-concept/feasibility data collection in the antenatal (before the onset of labour) study (PROSPEKT) is ongoing, funded by Wellcome Leap, to identify women with high-risk pregnancies who would benefit from an early intervention to reduce stillbirth. Our PPI focus group have reviewed all patient-focused documents for PROSPEKT and provided feedback regarding study protocol and dissemination of results. Our machine-learning (ML) model using pilot data correlated with pregnancy outcomes(1D) based on the recently developed near-miss criteria for stillbirth in global research: In Utero consensus9. For the goal of reducing ABI during childbirth, FetalSense should target labour monitoring to maximise the ability to detect placental compromise to escalate to prompt delivery. We propose the MERIT (Intrapartum fetoplacental monitoring with light) study, to understand the feasibility of using FetalSense in labour along with the currently available clinical monitoring tools. We will evaluate the quality of data and explore data preprocessing techniques required to make the raw data suitable for training ML models. We will also assess stakeholder perceptions regarding this technology.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know