Neonatal apnea and hypopnea prediction in infants with Robin sequence with neural additive models for time series

• DOI: 10.1371/journal.pdig.0000678
• Publication Date: 2024-12-13T18:37:55Z
• Authors: Julius Vetter; Kathleen Lim; Tjeerd M. H. Dijk...; Peter A. Dargaville; Oliver Kohlbacher; Jakob H. Macke; Christian F. Poets
• Keywords: Discriminative model; Hypopnea; Modalities

Neonatal apneas and hypopneas present a serious risk for healthy infant development. Treating these adverse events requires frequent manual stimulation by skilled personnel, which can lead to alarm fatigue. This study aims develop validate an interpretable model that predict hypopneas. Automatically predicting before they occur would enable the use of methods automatic intervention. We propose neural additive individual occurrences neonatal apnea hypopnea apply it physiological dataset from infants with Robin sequence at upper airway obstruction. The will be made publicly available together this study. Our proposed allows prediction hypopneas, achieving average AuROC 0.80 when discriminating segments polysomnography recordings starting 15 seconds onset control segments. Its nature makes inherently interpretable, allowed insights into how important given signal modality is patterns in are discriminative. For our problem sequence, prior irregularities breathing-related modalities as well decreases SpO 2 levels were especially presents step towards Together released dataset, has potential facilitate development application intervention clinical practice.