0956 Evaluation of a novel automated oxygen desaturation analysis

Abstract Introduction The oxygen desaturation index (ODI) quantifies the number of desaturations per hour sleep and is a valuable parameter in diagnosing sleep-related breathing disorders. While oximetry alone insufficient for an accurate diagnosis, it provides meaningful insights Obstructive Sleep Apnea (OSA) ODI has been shown to correlate with Apnea-Hypopnea Index (AHI). close relationship between other parameters especially interesting home-based studies and/or screening, which typically use limited set sensors (but almost always include oximetry) are becoming increasingly more prevalent. An automated tool analysis will thus be paramount clinician. Methods current study aims evaluate novel solution scoring, provided by Somnilog. We selected 20 overnight recordings, mix hospital- recordings from dataset Antwerp University hospital internal database collected several French centers. analyzed population (10 females, age 53.0±16.7 years old, BMI 28.8±4.8 kg/m²) had variety diagnoses, including non-apneic snoring central obstructive apnea. Each recording was also scored expert, using AASM guidelines 3% drop criterion. Results accurately detected events precision recall 0.97 0.96, respectively. reported intraclass correlation (ICC) 0.99 found expert. difference expert not significant lab- home-recorded patients (p=0.23). furthermore validated another well-used commercial package obtained similar performance metrics: 0.96 ICC 0.99. Conclusion demonstrates validity scoring. achieved near-perfect agreement experts, existing market. robust artifact-prone, thereby supporting increasing demand out-of-centre studies. Support (if any) n/a