Search Results

The Luna® G3 BPAP System is a Bi-level PAP (Bi-level Positive Airway Pressure) device designed for the treatment of adult Obstructive Sleep Apnea (OSA). The integrated humidification and warming of air from the flow generator device. These devices are intended for single-patient use by prescription in the home or hospital/ institutional environment on adult patients. It is to be used on patients > 66 lbs / 30 kg for whom CPAP therapy has been prescribed.

The Luna® G3 BPAP system is a microprocessor controlled, blower-based system that generates positive airway pressure to support treatment of obstructive sleep apnea. Its hardware design is identical to the previously cleared Luna® G3 BPAP 25A (K201620). The subject device includes two models, with different pressure ranges. They both have four therapy modes, which are CPAP, Spontaneous (S), Timed (T) and Spontaneous/Timed (S/T).

This 510(k) summary (K212263) describes a device modification to an existing Bi-level PAP system (Luna G3 BPAP System). The core of the submission revolves around demonstrating substantive equivalence to a predicate device (Luna G3 BPAP 25A, K201620) and referencing a device (Juno VPAP ST-A, K153061) for new functionalities.

It is crucial to understand that this document describes a device intended for treatment of Obstructive Sleep Apnea (OSA), providing positive airway pressure. It is not an AI/ML powered diagnostic device that would typically involve a multi-reader multi-case (MRMC) study or complex ground truth establishment from expert consensus or pathology, as might be seen for an imaging AI.

Therefore, many of the typical "acceptance criteria" and "study types" for AI/ML-powered devices, such as those related to standalone performance (e.g., sensitivity, specificity, AUC), MRMC studies, and expert ground truth establishment for a diagnostic output, are not applicable to this type of medical device submission.

Instead, the acceptance criteria and supporting studies are focused on engineering and performance characteristics of the medical device itself, ensuring it functions safely and effectively as intended and is equivalent to previously cleared devices.

Here's an analysis based on the provided document:

1. Table of Acceptance Criteria and Reported Device Performance

The document doesn't present a single aggregated table of "acceptance criteria" versus "reported performance" for a diagnostic output as would be seen for AI/ML diagnostic tools. Instead, it details specific performance tests and standards met, comparing the subject device's characteristics to its predicate and reference devices. The "acceptance criteria" are implied by adherence to recognized standards and demonstrated equivalence to the cleared predicate.

Here's an interpretation of relevant performance characteristics that serve as de facto acceptance criteria and their reported performance:

2. Sample Size Used for the Test Set and the Data Provenance

This is an engineering and performance validation study for a medical device, not a clinical study involving patient data for diagnostic accuracy.

• Sample Size: Not applicable in the traditional sense of patient data. The "test set" consists of the physical device units subjected to various engineering and performance tests (e.g., electrical safety, fluid dynamics, sound, software functionality). The document does not specify the number of physical units tested, but it would typically involve a small number of production or pre-production units.
• Data Provenance: Not applicable in the sense of country of origin of patient data. The tests are performed in a laboratory setting by the manufacturer, or by accredited third-party labs on behalf of the manufacturer, to verify product specifications and compliance with standards. The manufacturing location is indicated as Beijing, China (BMC Medical CO., LTD.).
• Retrospective or Prospective: Not applicable. These are engineering validation tests.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and the Qualifications of Those Experts

Not applicable. The "ground truth" for this type of device is defined by engineering specifications, validated performance against benchmarks (predicate and reference devices), and adherence to international recognized standards (e.g., ISO, IEC). There are no human "experts" establishing a clinical "ground truth" diagnosis for the device's output in the way a radiologist would for an imaging AI. The "experts" involved would be the engineers, quality assurance personnel, and regulatory affairs specialists who design, test, and approve the device.

4. Adjudication Method for the Test Set

Not applicable. There is no subjective interpretation or "adjudication" of results in the way there would be for a clinical diagnostic study with human readers. Test results are objective measurements against defined performance criteria and standards.

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

Not applicable. This is not an AI/ML-powered diagnostic device, nor is it designed to assist human readers in interpretation. It is a breathing therapy device. Therefore, an MRMC study is irrelevant to this submission.

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

This device does not have a "standalone" algorithmic component in the sense of a diagnostic AI assessing, for example, medical images. Its "algorithm" refers to the control software that manages the positive airway pressure delivery. The "standalone performance" refers to the device's ability to accurately deliver and control pressure, humidification, and other functions as per its specifications, which is what the various performance tests (e.g., static pressure, dynamic pressure, maximum flow rate, rise time, display accuracy) described under "Non-Clinical Performance Testing" are designed to evaluate. The results of these tests demonstrate the device's standalone performance in a lab setting.

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

The "ground truth" for this device's performance validation is based on:

• Engineering Specifications: The device's design parameters and target performance levels.
• International Standards: Adherence to recognized medical device standards (e.g., ISO 80601-2-70, ISO 80601-2-74, IEC 60601-1-2) which define acceptable ranges for various functional properties.
• Predicate Device Performance: The established, cleared performance characteristics of the Luna G3 BPAP 25A (K201620) and the relevant functionalities of the Juno VPAP ST-A (K153061). The current submission leverages the "substantial equivalence" argument, meaning the performance of the new device is compared to and validated as being equivalent to the legally marketed predicate.

8. The Sample Size for the Training Set

Not applicable. This is not an AI/ML model that is "trained" on a "training set" of data. The device's internal control software/algorithm is based on fixed programming, not machine learning.

9. How the Ground Truth for the Training Set was Established

Not applicable for the same reason as point 8. The "ground truth" for the device's design and programming comes from established physiological principles of breathing support and engineering principles for mechanical ventilation devices.

Ask a specific question about this device