Related Experiment Video
Updated: Jan 8, 2026
19:32
Preterm EEG: A Multimodal Neurophysiological Protocol
Published on: February 18, 2012
28.9K
Unobtrusive cot side sleep stage classification in preterm infants using ultra-wideband radar
Emad Arasteh1,2, Eline R de Groot1, Demi van den Ende1
1Department of Neonatology, University Medical Center Utrecht, Wilhelmina Children's Hospital, Utrecht, Netherlands.
Frontiers in Sleep
|December 22, 2025
Summary
Ultra-wideband (UWB) radar shows promise for monitoring sleep stages in preterm infants within the neonatal intensive care unit (NICU). This non-obtrusive method achieved 81% accuracy in classifying active and quiet sleep, aiding infant development assessment.
Area of Science:
- Neonatal Medicine
- Biomedical Engineering
- Sleep Science
Background:
- Sleep is crucial for preterm infant development.
- Continuous, non-obtrusive sleep monitoring in neonatal intensive care units (NICUs) presents significant challenges.
Purpose of the Study:
- To evaluate the feasibility of using ultra-wideband (UWB) radar for sleep stage classification in preterm infants.
- To determine if UWB radar can accurately differentiate between active and quiet sleep in this vulnerable population.
Main Methods:
- Ten preterm infants (29-34 weeks postmenstrual age) in the NICU were monitored using UWB radar and video recordings for sleep assessment.
- UWB radar captured infant motion, from which 48 features related to body and breathing movements were extracted.
- Six machine learning classifiers were compared for their effectiveness in classifying sleep stages based on UWB radar data.
Main Results:
- The Adaptive Boosting (AdaBoost) classifier demonstrated the highest performance.
- AdaBoost achieved a balanced accuracy of 81% and an Area Under the Curve (AUC-ROC) of 0.82 in 10-fold cross-validation.
- The classifier effectively distinguished between active and quiet sleep stages.
Conclusions:
- UWB radar technology, coupled with the AdaBoost classifier, offers a promising non-obtrusive approach for assessing sleep stages in very preterm infants.
- This method has the potential to improve developmental monitoring for infants in the NICU.
