FDA Clears HoneyNaps Somnum V3.0 AI Software for Automated Sleep Apnea Classification (97% Agreement)
Tags AI · Enterprise
The FDA granted 510(k) clearance to HoneyNaps' Somnum V3.0 AI software for polysomnography analysis. The tool automatically detects and classifies obstructive, central, and mixed sleep apnea events with >97% overall percent agreement across categories. It provides event-level classification rather than composite indices, supporting clinical scoring. HoneyNaps plans future 510(k) submissions adding hypoxic burden, arousal burden, and ventilatory burden biomarkers for severity assessment.
Technical significance
Event-level apnea subtyping (OSA/CSA/MSA) via multi-channel biosignal analysis addresses a known limitation of manual scoring — differentiating central vs. obstructive events in complex cases. The 97% agreement meets clinical-grade thresholds. Integrating hypoxic/arousal/ventilatory burden biomarkers could enable personalized severity stratification beyond AHI (Apnea-Hypopnea Index). FDA clearance for AI-assisted PSG scoring expands the reimbursable clinical workflow for sleep labs facing technician shortages.