(319 days)
The Galapagos app is intended for patients who are prescribed a compatible ResMed S10 platform device to simulate therapy prior to using their device with their prescribed settings. It is an optional software accessory to allow patients to acclimate to their therapy device.
Galapagos is a mobile interface where the patient can control the device through pre-determined, scaled inspiratory pressures to help them adjust to pressure and mask fit prior to starting their prescribed therapy. In addition, Galapagos may also be used as a communication pathway using the mobile Bluetooth connection to the compatible flow generator in order to send and receive data.
The provided text is a 510(k) summary for the device "Galapagos," a mobile application. It details the device's indications for use, its comparison to a predicate device (Monte Carlo), and the testing performed to demonstrate substantial equivalence.
However, the document states: "Clinical tests were not required to demonstrate the safety and effectiveness of Galapagos."
This crucial statement means that the acceptance criteria and study proving the device meets those criteria were not based on clinical performance studies like those typically conducted for AI/ML medical devices (e.g., MRMC studies, standalone performance with reader ground truth). Instead, the assessment for this device relied on non-clinical verification and validation testing to demonstrate that it met its intended performance requirements and was substantially equivalent to a previously cleared predicate device.
Therefore, I cannot provide the information requested in points 1-9 as they pertain to clinical performance studies and AI/ML model validation, which were explicitly not conducted for Galapagos. The acceptance criteria and "proof" of meeting them for Galapagos are primarily based on software verification and validation, and bench testing, rather than human-in-the-loop or standalone AI performance.
The document highlights the following non-clinical testing for Galapagos:
- Software verification and validation
- Predicate testing (which involved bench tests against nonfunctional requirements and end-to-end functional testing, comparing Galapagos to the Monte Carlo device).
The "acceptance criteria" here is therefore the successful completion of these non-clinical tests and the demonstration of substantial equivalence based on differences not impacting safety or efficacy.
To directly address your request based on the provided text, but acknowledging the limitations:
-
A table of acceptance criteria and the reported device performance:
- Acceptance Criteria (Implied from the document): The device must successfully pass software verification and validation, and non-clinical performance bench tests including end-to-end functional testing and predicate testing (demonstrating substantial equivalence to Monte Carlo without raising new questions of safety or efficacy).
- Reported Device Performance: "Non-clinical verification and validation testing completed for Galapagos demonstrated that the device met all intended performance requirements." And "The predicate testing conducted with Galapagos to the predicate Monte Carlo device (K160836) demonstrate substantial equivalence to the technology and operating principle of the devices."
-
Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective): Not applicable, as clinical data was not used for performance evaluation. The "test set" would refer to the scenarios and inputs used during software/bench testing, the details of which are not provided.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts: Not applicable. Ground truth for clinical performance was not established due to the absence of clinical testing.
-
Adjudication method (e.g. 2+1, 3+1, none) for the test set: Not applicable.
-
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 MRMC study was done, as clinical tests were not required. The device is a "mobile interface where the patient can control the device" and an "optional software accessory to allow patients to acclimate to their therapy device," not an AI/ML diagnostic or assistive tool for human readers.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done: Not applicable in the context of typical AI/ML standalone performance studies. Bench and software tests were done, but these are not the same as standalone diagnostic performance of an AI algorithm on patient data.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc): Not applicable for clinical ground truth. The "ground truth" for the non-clinical testing would be the expected functional behavior and performance characteristics derived from design specifications and predicate device equivalence.
-
The sample size for the training set: Not applicable, as this device does not appear to be an AI/ML model that learns from a training set of data. It's a software application controlling a medical device.
-
How the ground truth for the training set was established: Not applicable.
§ 868.5905 Noncontinuous ventilator (IPPB).
(a)
Identification. A noncontinuous ventilator (intermittent positive pressure breathing-IPPB) is a device intended to deliver intermittently an aerosol to a patient's lungs or to assist a patient's breathing.(b)
Classification. Class II (performance standards).