ApneaSenseby SomniAI LLC

Research & validation

The methodology, the numbers, and the code — checkable.

These are SomniAI LLC's task-specific validation results against PSG reference annotations. Three technical preprints, three open-source repositories. Check the task, unit, denominator and limitations before using any number.

The validation set

The snore-segment and breathing-window classification tasks were evaluated against reference annotations from in-lab and ambulatory polysomnography (PSG): 80 paired PSG nights across 40 participants — 10 in-lab PSG and 70 ambulatory PSG nights with a nasal-airflow cannula, with smartphone audio captured simultaneously.

Illustrative time-aligned acoustic and PSG reference tracks used to explain validation alignment.
Illustrative A schematic of time-aligned acoustic input and PSG reference annotations — not plotted study data. The published metrics below apply to their stated 1-second segment or 200-second window tasks.

The complete system — one two-stage cascade

The production model is a two-stage on-device audio cascade: Stage-1 snore classification feeds Stage-2 breathing-window classification, with Coordinate-Attention 1D as the Stage-2 method. The combined system is documented in a Research Square preprint with a citable DOI; it is not peer-reviewed. Read the preprint → · Google Scholar →

How it's built — baseline → attention → production

The same work, broken into the three steps it was built from — a cascade, its attention-based Stage-2 classifier, then that classifier compressed for the device. Each is its own Zenodo preprint with an MIT-licensed code companion.

Step 1 · baseline & cascade

Two CNN baselines + the cascaded two-stage pipeline

The baseline snore and apnea CNNs, with multi-seed bootstrap validation — and the Stage-1→Stage-2 cascade that the next two build on.

94.29%snore accuracy (93.60–95.02)
83.82%apnea baseline accuracy
0.983snore AUC-ROC
Read the paper (PDF) → Run the code (GitHub) → Cite — Zenodo DOI →
Step 2 · attention (Stage-2)

Coordinate-Attention 1D — the Stage-2 classifier

Replaces the baseline apnea CNN above as the cascade’s Stage-2 classifier: a 14,001-parameter attention model that keeps temporal position — a 93.2% parameter cut, with accuracy preserved or improved.

87.14%accuracy (85.14–89.68)
14,001parameters
93.2%param reduction vs baseline
Read the paper (PDF) → Run the code (GitHub) → Cite — Zenodo DOI →
Step 3 · compression → production

Compression → the on-device production model

Compresses the Stage-2 model above to production size: INT8 quantization-aware training + 50% structured pruning + CoreML. On this dataset, compression raises test accuracy.

88.49%breathing-window accuracy (INT8 + pruned)
56.4 KB9,416 INT8 params
0.064 msinference · Apple M2 Neural Engine
Read the paper (PDF) → Run the code (GitHub) → Cite — Zenodo DOI →

What we publish, and what the patent covers

The preprints and code disclose the research method — the cascaded two-stage architecture over a compact 200×3 @ 1 Hz representation, the Coordinate-Attention 1D formulation, the quantization and pruning protocol, and the evaluation procedure. The implementation can be inspected and rerun on a buyer's own corpus. The original validation corpus is not distributed, so an exact independent reproduction on the same data is not possible.

Data availability

The validation corpus — the 80 paired PSG nights and the simultaneous smartphone audio — is not publicly distributed. Participants consented to internal validation, not redistribution, so the raw recordings and labels stay private. What's public is the method and code: each preprint above has an MIT-licensed companion repository — algorithm framework and training code, no data — so teams can run the same evaluation procedure on their own corpus.

Limitations we hold to

  • Acoustic proxy. We detect sound, not airflow or blood oxygen — PSG and continuous SpO₂ measure the physiological event more directly. We don't claim to beat them.
  • Screening, not diagnosis. ApneaSense is not a medical device and is not FDA-cleared.
  • Precision-first. Tuned to avoid false positives, so it under-counts — read the rate as a conservative lower bound.
  • Sample. Recruited with sleep-breathing complaints; generalization across BMI and healthy cohorts is ongoing.

Citing this work

Authored at SomniAI LLC by the inventor on the pending U.S. patent (PAT-001). Each paper is published on Zenodo with a citable DOI — linked under each paper above; the combined cascade paper is on Research Square. Full profile on Google Scholar and ORCID.

Want to evaluate it on your own data? See how the SDK works or scope a focused pilot.