Search Results

SleepMap system is intended for use as an aid for physicians in the diagnosis of sleep and respiratory related disorders in adult patients (aged 18 and up), who have been prescribed a sleep study by their doctor.

SleepMap is a software-only medical device to be used under the supervision of a clinician to analyze physiological signals and automatically score sleep study recordings, including the staging of sleep, detection of arousals, leg movements, desaturations, obstructive apneas and obstructive hypopneas.

Respiratory event subtypes (central and mixed apneas; central hypopneas), RERA, Cheyne Stokes Breathing, Snoring Events, and Arrhythmia Events are not automatically detected and must be manually marked within records.

All automatically scored stages and detected events can be manually marked or edited within records during review.

All automatically scored stages, detected events and physiological signals which are retrieved, analyzed, displayed, and summarized are subject to verification by a qualified clinician.

Onera SleepMap is a software-only medical device that analyses previously recorded physiological signals obtained during sleep through a polysomnography sleep test with the Onera STS I device. SleepMap can analyze at-home or in-lab sleep study recordings of adult patients.

Automated algorithms are applied to the raw input signals (read from a measurement file in the EDF file format originating from the Onera STS I device) in order to derive additional signals:

• a heartrate (HR) signal from the raw ECG signal. The heartrate algorithm does not include any automated arrhythmia analysis.
• a quantized position signal from the raw continuous value position signal.

Additionally, multiple algorithms are used to interpret the raw and derived signals by classifying sleep stages, sleep events and artifacts. The software automates recognition of the following sleep events: obstructive apnea and obstructive hypopnea events, arousal events, desaturation events, leg movement and PLM events.

The Onera SleepMap contains the following algorithms:

• Sleep staging algorithm, a deep learning model (AI-model) which classifies sleep stages, based on EEG, EOG and EMG inputs.
• Arousals algorithm, a Convolutional Neural Network (AI-model) which predicts arousals, based on sleep stages, EEG, EOG, EMG, ECG and nasal pressure inputs.
• Desaturation algorithm, a rule-based algorithm which detects events of minimum 3% or 4% oxygenation drop based on sleep stages and SpO2 signal inputs.
• Apnea (obstructive) detection, a rule-based algorithm which detects ≥90% nasal pressure drops, based on sleep stages and nasal pressure signal inputs.
• Hypopneas (obstructive) detection, a rule-based algorithm which detects ≥30% nasal pressure drops based on sleep stages, nasal pressure signal, arousal events, and desaturation event inputs.
• Leg Movement algorithm, a rule-based algorithm which detects (repetitive) EMG amplitude increasements, based on sleep stages, EMG signal and respiratory event inputs.
• Artifact algorithms, rule-based algorithms which detect artifacts, based on SpO2, EEG, EOG, EMG and heart rate inputs.

Additionally, clinical users can manually annotate: respiratory event subtypes (central and mixed apneas and central hypopneas), RERA, Cheyne Stokes Breathing, Snoring Events, and Arrhythmia Events.

The raw signals, derived signals and all automated analysis results (annotations) must be visually inspected and reviewed by sleep analysts and physicians prior to the results being integrated into a sleep study report.

SleepMap calculates aggregated metrics and indexes on the set of annotations resulting from the sleep analyst or physician review and integrates these into a technical sleep report that can be previewed.

The technical sleep study report summarizes the sleep stage annotations in a hypnogram, provides the aggregated metrics and indexes, and the technician notes into a PDF document which is the main output of SleepMap. The technical sleep study report is transferred to the Onera Digital Health Platform for storage and is used by the physician to diagnose the sleep disorder.

Here's a breakdown of the acceptance criteria and study details for the Onera SleepMap, based on the provided FDA 510(k) clearance letter:

Onera SleepMap Acceptance Criteria and Study Summary

The Onera SleepMap is a software-only medical device intended as an aid for physicians in the diagnosis of sleep and respiratory related disorders in adult patients (18 and up) who have been prescribed a sleep study. It analyzes physiological signals and automatically scores sleep study recordings for sleep staging and the detection of arousals, leg movements, desaturations, obstructive apneas, and obstructive hypopneas.

1. Acceptance Criteria and Reported Device Performance

The acceptance criteria for the Onera SleepMap's automated scoring algorithms are implicitly defined by the reported performance metrics (Positive Agreement, Negative Agreement, Overall Agreement, False Discovery Rate) against a 2/3 majority ground truth from expert scorers. The performance metrics are presented with 95% bootstrap confidence intervals. The device is deemed acceptable if these metrics demonstrate sufficient agreement with human expert consensus, indicating it can serve as a reliable aid for physicians.

2. Sample size used for the test set and the data provenance

• Automated Scoring Service (Sleep Staging & Events):
  • Sample Size: 98 PSG night studies from 98 unique patients.
  • Data Provenance:
    • N=72 studies from the United States (home environment), acquired August 2023 to August 2025.
    • N=26 studies from Germany (clinic environment), acquired September 2022 to April 2024.
• Heartrate (HR) Algorithm:
  • Sample Size: 249 PSG night studies from 215 unique patients. The validation used 5-minute time intervals from these studies.
  • Data Provenance:
    • N=171 studies from the United States (home environment), acquired August 2023 to August 2025.
    • N=78 studies from Germany (clinic environment), acquired September 2022 to April 2024.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• Automated Scoring Service (Sleep Staging & Events):
  • Number of Experts: 3
  • Qualifications: Independent, US-based certified sleep professionals.
• Heartrate (HR) Algorithm:
  • Number of Experts: 2 independent cardiac technologists initially for manual annotation, and an expert board-certified cardiologist for adjudication.
  • Qualifications: Independent cardiac technologists and an expert board-certified cardiologist from Florida, USA.

4. Adjudication method for the test set

• Automated Scoring Service (Sleep Staging & Events): A 2 out of 3 (2/3) majority scoring was constructed as the ground truth reference. Epochs where a 2/3 majority agreement was not reached (n=3282 for sleep staging) were excluded from validation.
• Heartrate (HR) Algorithm: Disagreements in annotations between the 2 cardiac technologists were adjudicated by an expert board-certified cardiologist.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

The provided document does not indicate that an MRMC comparative effectiveness study was performed to evaluate human readers' improvement with AI assistance. The study focuses on the standalone performance of the AI algorithms against expert consensus.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

Yes, a standalone performance study was done. The reported performance metrics (Positive Agreement, Negative Agreement, Overall Agreement, False Discovery Rate, and absolute error for HR) describe the Onera SleepMap algorithms' performance independently against the established ground truth. The device is designed for use "under the supervision of a clinician," and all automatically scored stages and events "are subject to verification by a qualified clinician," indicating that the reported standalone performance is intended as an aid for, not a replacement of, human review.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

• Automated Scoring Service (Sleep Staging & Events): Expert consensus, specifically a 2/3 majority agreement from 3 independent, US-based certified sleep professionals, in compliance with AASM scoring guidelines.
• Heartrate (HR) Algorithm: Expert consensus derived from manual annotations by 2 independent cardiac technologists, with disagreements adjudicated by an expert board-certified cardiologist.

8. The sample size for the training set

The document does not specify the sample size used for the training set for any of the algorithms. It only describes the validation data sets.

9. How the ground truth for the training set was established

The document does not specify how the ground truth for the training set was established. It only details the ground truth establishment for the validation data set.

Ask a specific question about this device