Novel single-lead electrocardiogram-derived respiratory rate estimation in children with obstructive sleep apnea

<bold>Background:</bold> Respiratory rate (RR) monitoring is critical in the assessment of obstructive sleep apnea, diagnosed in ~20 million children each year. RR is estimated using inductive plethysmography (RIP) which utilizes specialized equipment and is cumbersome and restrictive. <bold>Objective:</bold> Investigate the clinical feasibility of RR estimation using single-lead body-surface electrocardiograms (ECG) in pediatric subjects. <bold>Methods:</bold> Data were obtained from 40 pediatric patients (n=20 male, 20 female) aged 0-20 years undergoing a polysomnographic sleep study, from the National Sleep Research Resource repository. 5-minute segments with stable ECG and RIP-estimated RRs were identified. A novel algorithm was used to estimate RRs from the ECG, which were compared with those in the repository obtained using RIP. <bold>Results:</bold> RR acquired during polysomnography (RIP, red) and ECG-derived RR (ECG, blue) from one subject are depicted in Figure 1A. The algorithm-derived RR rapidly tracked the RIP-determined RR in the spontaneously breathing child. A close linear correlation was observed between the reference and algorithmic RRs across the entire cohort (Figure 1B). The mean absolute RR estimation error was 0.8±0.8 breaths/min (Figure 1C), while the mean relative RR estimation error was 4.3±4.5%. <fig><object-id>erj;64/suppl_68/PA1354/F1</object-id><object-id>F1</object-id><object-id>F1</object-id><graphic></graphic></fig> <bold>Conclusion:</bold> Our novel, single-lead ECG-derived RR estimation demonstrates high fidelity and precision in pediatric subjects.