Acta Paediatr. 2026 Sep 6. doi: 10.1111/apa.70744. Online ahead of print.
ABSTRACT
AIM: Respiratory rate is a key vital sign in paediatric assessment, but accurate measurement in emergency settings is difficult because conventional methods (manual counting, capnography) can be unreliable or impractical. Respiratory rate is frequently omitted in busy emergency departments.
METHODS: We developed and evaluated a non-contact thermal imaging system that automatically measures respiratory rate and the inspiratory-to-expiratory ratio in paediatric patients. Thermal imaging detects temperature changes around the nostrils produced by inhaling cooler air and exhaling warmer air. Respiratory signals were obtained from 13 paediatric patients (14 datasets, aged 3-13 years) using an uncooled thermal camera. Automated image processing extracted respiratory waveforms and identified respiratory cycles and their inspiratory and expiratory phases using autocorrelation and peak detection.
RESULTS: Clear respiratory signals were obtained for all patients, including those wearing masks. Several respiratory rate values were outside published reference percentiles, but all participants were clinically stable with values consistent with physiological variability. The mean inspiratory-to-expiratory ratio was 1.20 ± 0.23.
CONCLUSION: These findings demonstrate the feasibility of a fully non-contact thermal system for automatic measurement of respiratory rate and the inspiratory-to-expiratory ratios during mask use. Further validation in larger cohorts with comparison to a gold standard is required before clinical adoption.
PMID:42701864 | DOI:10.1111/apa.70744
