Oximetry Analysis
17-Metric SpO₂ & Heart Rate Framework
AirwayLab's oximetry pipeline provides a comprehensive analysis of overnight pulse oximetry data from Viatom/Checkme O2 Max wrist oximeters. Seventeen metrics across five categories give you a detailed picture of oxygenation, cardiac response, and their interaction throughout the night.
How it works
- 01CSV import — Upload the CSV export from your Viatom/Checkme O2 Max alongside your ResMed SD card data. The parser extracts SpO₂ and heart rate time series with their timestamps.
- 02Artefact correction — Double-tracking artefacts (where the oximeter reports two readings for the same event) are detected and corrected to prevent inflated event counts.
- 03Multi-metric analysis — Seventeen metrics are computed across five categories: oxygen desaturation, heart rate surges, coupled events, summary statistics, and half-night comparisons.
- 04Clinical integration — Results are presented alongside PAP flow analysis data, enabling you to see how flow limitation events correlate with oxygen desaturations and heart rate responses.
The 17 metrics
Oxygen Desaturation
ODI-3%
Oxygen Desaturation Index at 3% threshold — the number of times per hour SpO₂ drops by ≥3% from baseline. The most widely used clinical desaturation metric.
ODI-4%
Oxygen Desaturation Index at 4% threshold — a stricter metric used in some clinical guidelines. Correlates more strongly with cardiovascular outcomes.
T<90%
Percentage of total recording time spent with SpO₂ below 90%. Extended time below 90% is associated with increased cardiovascular risk.
T<94%
Percentage of recording time with SpO₂ below 94%. A more sensitive threshold that captures moderate hypoxemia often missed by the 90% cutoff.
Heart Rate Analysis
Clinical HR Surges (8 bpm)
Heart rate increases >8 bpm above a 30-second trailing baseline. These surges often accompany respiratory arousals and correlate with sleep fragmentation.
Clinical HR Surges (10 bpm)
Heart rate increases >10 bpm above 30-second baseline — a stricter threshold identifying more significant autonomic activations.
Clinical HR Surges (12 bpm)
Heart rate increases >12 bpm above 30-second baseline — captures only larger autonomic responses, reducing false positives from normal HR variability.
Clinical HR Surges (15 bpm)
Heart rate increases >15 bpm above 30-second baseline — the strictest clinical threshold, identifying the most pronounced arousal-related heart rate spikes.
Rolling Mean HR Surges (10 bpm)
Heart rate increases >10 bpm above a 5-minute rolling mean. This longer baseline captures surges relative to the prevailing heart rate trend.
Rolling Mean HR Surges (15 bpm)
Heart rate increases >15 bpm above 5-minute rolling mean — the stricter version of the rolling mean surge metric, identifying only the most significant deviations.
Coupled Events
Coupled ODI + HR Events
Simultaneous SpO₂ desaturation and heart rate surge within a time window. These coupled events strongly suggest respiratory arousal — the desaturation reflects the apnea/hypopnea and the heart rate surge reflects the arousal response.
Summary Statistics
Mean & Min SpO₂
Mean and minimum oxygen saturation across the recording. Minimum SpO₂ captures the deepest desaturation event.
Mean HR & HR Variability
Mean heart rate and standard deviation. Elevated mean HR may indicate poor sleep quality or increased sympathetic tone. HR variability reflects autonomic regulation.
Half-Night Comparison
H1 vs H2 Metrics
All desaturation and heart rate metrics are computed separately for the first and second halves of the recording. Differences between halves can reveal positional effects (if you change position during the night) or REM-related patterns (REM sleep is concentrated in the second half of the night and is when the airway is most collapsible).
Compatible devices
AirwayLab currently supports CSV exports from the Viatom/Checkme O2 Max wrist pulse oximeter. This is a popular choice in the sleep apnea community for its comfort, reliability, and ability to export second-by-second data.
Oximetry data is entirely optional — all three PAP flow analysis engines (Glasgow Index, WAT, NED) work with SD card data alone. The oximetry pipeline activates only when you provide oximetry CSVs alongside your SD card upload.