Search Results

The Sunrise Air is a non-invasive home care aid in the evaluation of obstructive sleep apnea (OSA) in patients 18 years and older with suspicions of sleep breathing disorders.

The Sunrise Air consists of the Sunrise software (v1.28.00), which analyzes data from one of three compatible sensors (Sunrise sensor 1, Sunrise sensor 2, or Sunrise Air) placed on the patient's chin. Sunrise sensor 1 was approved through DEN210015, while Sunrise sensor 2 was cleared through K222262. The current version of the Sunrise device introduces a new sensor, Sunrise Air. The Sunrise device is intended to detect respiratory events, identify sleep stages and position, and generate key sleep parameters—such as the apnea-hypopnea index ("Sunrise AHI") and positional states classifications. The collected data is compiled into a report for further interpretation by a healthcare provider.

The provided FDA 510(k) clearance letter for the Sunrise Air device primarily focuses on demonstrating substantial equivalence to a predicate device, rather than detailing a comprehensive clinical study to prove the device meets specific acceptance criteria for its claimed indications.

The document highlights bench testing for technical equivalence, but lacks the detailed clinical study information typically provided for direct performance claims against established ground truth. Specifically, it states that "No modifications have been made to the Sunrise algorithm used to generate sleep parameters," and that a "validation study of SpO₂ and pulse rate accuracy for the subject device was conducted using raw PPG data acquired during the clinical validation for the Sunrise sensor 2 (K222262)." This suggests reliance on prior clearances for core algorithm performance and a specific re-validation for only the PPG data processing change.

Therefore, many of the requested details about acceptance criteria, clinical study design, and ground truth establishment for the overall device performance (e.g., AHI calculation, OSA evaluation) are not explicitly present in this summary.

Given the information in the provided document, here's what can be extracted and inferred:

1. A table of acceptance criteria and the reported device performance:

Based on the information provided, the "acceptance criteria" are implied by the comparisons to the predicate and reference devices, and some specific performance metrics are given for SpO2 and pulse rate. The primary acceptance criterion for the device's main function (evaluation of OSA via AHI) is that "No modifications have been made to the Sunrise algorithm used to generate sleep parameters," implying continued equivalence to the predicate's performance.

Study that Proves the Device Meets Acceptance Criteria:

The document describes a combination of bench testing and reliance on prior clinical validation for specific components. There isn't a single, new "study" designed to prove the overall device meets a set of clinical acceptance criteria for OSA evaluation, but rather, individual tests to establish equivalence of components or re-validate specific algorithm changes.

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

• Overall Device (for AHI/OSA evaluation): Not explicitly stated for a new study. The document states that "No modifications have been made to the Sunrise algorithm used to generate sleep parameters." This implies reliance on the clinical validation data from the predicate device (Sunrise K222262). The original K222262 submission would contain this information.
• SpO₂ and Pulse Rate Accuracy:
  • Sample Size: Not explicitly stated. The study was conducted using "raw PPG data acquired during the clinical validation for the Sunrise sensor 2 (K222262)." The sample size for that original clinical validation would be the relevant number.
  • Data Provenance: Retrospective, as it used data from a previous clinical validation study (for Sunrise sensor 2, cleared under K222262). The country/region of origin of this data is not specified in this document.

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

• Overall Device (for AHI/OSA evaluation): Not specified in this document, as the core algorithm relies on prior validation. For the original K222262 clearance, ground truth would typically be established by a consensus of sleep experts (e.g., board-certified sleep physicians or registered polysomnographic technologists (RPSGTs)).
• SpO₂ and Pulse Rate: Ground truth for these parameters is typically established through a co-oximeter or arterial blood gas analysis, not necessarily by "experts" in the human sense, but by a gold-standard measurement device.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

• Not specified within this 510(k) summary for any new studies. For the original clinical validation of the AHI algorithm, an adjudication method (such as independent scoring by multiple qualified technologists/physicians with consensus or a tie-breaker) would typically be employed for the polysomnography (PSG) ground truth.

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:

• No, an MRMC comparative effectiveness study involving human readers and AI assistance is not described in this document. The device is for "aiding in the evaluation" and generates parameters; it is not presented as an AI-assisted diagnostic tool for human interpretation improvement in this summary.

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

• Yes, the device's capability to "detect respiratory events, identify sleep stages and position, and generate key sleep parameters" and a "Sunrise AHI" implies a standalone algorithmic performance in generating these outputs from the sensor data. The statement "No modifications have been made to the Sunrise algorithm used to generate sleep parameters" means that the standalone performance of the algorithm itself is considered validated based on its prior clearance. The SpO₂ and pulse rate accuracy also represent standalone algorithm performance.

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

• Overall Device (for AHI/OSA evaluation): Not explicitly stated, but for sleep apnea diagnostic devices, the ground truth is overwhelmingly polysomnography (PSG) scored by qualified experts (e.g., according to AASM guidelines). This would have been the ground truth for the predicate device's (K222262) clearance.
• SpO₂ and Pulse Rate: The ground truth for SpO₂ accuracy is typically established using a reference pulse oximeter or co-oximeter (invasive arterial blood gas analysis may be used for a subset of the data if required for the specific accuracy claims and range). For pulse rate, a simultaneous ECG or the reference oximeter's heart rate measurement.

8. The sample size for the training set:

• Not specified in this document. As the core algorithm is unchanged from the predicate, its training data would have been part of the K222262 submission.
• The document mentions "cloud-based algorithm (Sunrise PPG algorithm)" as a change for PPG data processing, but it does not specify the training set size for this particular component, only that its validation was done on existing test data.

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

• Not specified in this document, as the core algorithm leverages prior clearance. For the predicate device, ground truth for training data would have broadly been established in the same manner as the test set: expert-scored polysomnography (PSG) data. However, the specific details (e.g., single expert vs. consensus) are not provided here.

Ask a specific question about this device