1. Validation and certification
- Is there independent evidence, such as AASM Autoscoring Certification or peer-reviewed evaluation, rather than only vendor-reported figures? (AASM (opens in a new tab))
- Was it tested on a population similar to your patients (age, sleep disorders, recording montage)?
- Does the vendor report performance per stage, especially N1, not just overall accuracy?
- What is its regulatory status in your jurisdiction, and for what intended use?
2. Review and correction tools
- Per-epoch confidence or probabilities, so reviewers can prioritise uncertain epochs.
- Queues or filters for low-confidence, N1, transition and unreviewed epochs.
- Fast correction with notes, undo and redo.
- An audit trail that keeps the original model output and every change.
- A sign-off step that requires every epoch to be reviewed.
3. Data handling
- Where recordings are processed and stored: on your workstation or on a vendor's servers.
- Whether any data, logs or telemetry leave the machine.
- How access, encryption, backups and retention are handled. See local vs cloud sleep scoring.
4. Formats and outputs
- Which recording formats it imports (for example EDF and EDF+), and whether it checks channel labels, sampling rates and units.
- Per-epoch exports (such as CSV) and a readable report (such as PDF).
- Clear definitions for summary measures, and how lights-off and lights-on times are handled.
- Scope: does it stage sleep only, or also score respiratory events, arousals and limb movements?
5. Licensing, training and support
- Licensing of the software and any pretrained models for your intended use.
- Installation, configuration and technician training.
- Written support terms and a documented handover.
How InEpoch answers this checklist today
| Item | InEpoch today |
|---|---|
| AASM Autoscoring Certification | No |
| Regulatory clearance (FDA, CDSCO or other) | None claimed |
| Independent clinical validation of the full application | Not yet; we run local pilots against your scorers |
| Staging engine | YASA's published, peer-reviewed classifier |
| Per-epoch probabilities and review queues | Yes |
| Audit trail, undo/redo and sign-off | Yes |
| Processing location | Local workstation; no cloud upload required for analysis |
| Formats and exports | EDF/EDF+ import; per-epoch CSV and PDF report |
| Scope | Sleep staging only; no respiratory, arousal or limb-movement scoring |
| Installation, training, support | Yes: the core of the InEpoch service |
We publish this so you can compare like with like. To discuss a pilot, contact InEpoch.
Sources
- AASM Autoscoring Certification Program (opens in a new tab)
- Goldstein CA et al. Artificial intelligence in sleep medicine: an AASM position statement. J Clin Sleep Med. 2020;16(4):605–607 (opens in a new tab)
- Vallat R, Walker MP. An open-source, high-performance tool for automated sleep staging. eLife. 2021;10:e70092 (opens in a new tab)