(98 days)
The VISIONAIR™ system is a software application intended to be used with third-party endoscopic systems in the measurement of the nasal respiratory airway. The VISIONAIR™ system measures the nasal respiratory airway from the endoscopic images taken in the region of the internal nasal valve (INV) and nasal cavum (NC).
The VISIONAIR™ software can be utilized to automatically measure the cross-section area of the Internal Nasal Valve (INV) and Nasal Cavum (NV) and measure the nasal respiratory airway in this region of the anatomy. The VISIONAIR™ system consists of the following components: - The VISIONAIR™ algorithm which performs the Internal Nasal Valve and Nasal Cavum cross-section area segmentations. - The VISIONAIR™ Graphical User Interface (GUI) used for data entry, view the endoscopic image from third party endoscopes and display the data analysis to the user. - A smart device such as tablet or laptop which runs on Windows 10 or a later operating system with the VISIONAIR™ application installed. - A cloud service that runs in the background and can be activated by the user when a particular dataset for a case is desired to securely and anonymously be stored to the Cloud Server (REAI). - USB memory used to encrypt and anonymize the patient information, whether the data is stored locally or to the cloud, and stores the credits needed to activate the VISIONAIR™ software for each case. The VISIONAIR™ application interfaces with third-party endoscopic systems via the ports located on the smart device. The smart device ports enable the third-party systems endoscopic video display to be streamed on the VISIONAIR™ application endoscopic video display. In this 510(k) submission, a FDAcleared endoscope (K970247) was selected as the reference device to support the scientific methodology. The VISIONAIR™ application automatically analyzes the endoscopic images using its trained AI algorithm to measure the nasal valve and nasal cavum surface areas. The VISIONAIR™ application also provides a database file system to manage the data and interface securely and anonymously with the cloud server via the REAl module.
The provided FDA 510(k) summary for the VISIONAIR™ system offers details on its intended use and comparison to a predicate device, as well as a list of non-clinical tests performed. However, it does not explicitly state specific acceptance criteria (e.g., minimum accuracy thresholds) or present the detailed results of a study designed to prove the device meets those criteria with statistical significance.
Instead, it lists tests performed, implying that these tests confirmed design specifications and functionality. The "Substantial Equivalence Table" focuses on comparing attributes to a predicate device and concluding that differences do not raise new safety or effectiveness concerns.
Therefore, many of the requested details about acceptance criteria, specific study results, sample sizes, ground truth establishment, expert qualifications, and MRMC studies are not present in the provided text.
Based on the available information, here's what can be extracted and what is missing:
Acceptance Criteria and Reported Device Performance
The document does not explicitly state numerical acceptance criteria (e.g., "accuracy > 90%"). Instead, the "Non-clinical Performance Data" section lists various tests performed to ensure the device functions according to design specifications and for substantial equivalence in terms of safety and effectiveness. The "reported device performance" is largely implied by the statement that these tests were "performed" and that the device is deemed "substantially equivalent."
| Acceptance Criteria Category (Implied) | Reported Device Performance (Implied from document) |
|---|---|
| System Functionality | Confirmed successful operation across various components: |
- System Level Test: Confirmation of Windows OS, processor, RAM, ports, wireless connectivity.
- System Interface and Connectivity Test: Confirmation of application to USB device (cloud key, credits) and connections to other devices.
- VISIONAIR™ Application Test: Confirmed connectivity to external endoscopes, cloud server, successful launch, and interaction tests.
- Endoscopic Display Test: Endoscopic view verification of image capture, video recording, and other functions. |
| Data Management & Security | Confirmed successful execution of data handling and security features: - Patient Database Verification Test: Confirmation of data stored, anatomical marking, and successful encryption/decryption of the database.
- Report Generation Test: Confirmation of successful report generation in pdf, csv, and other formats. |
| AI (Segmentation) Accuracy | Evidence of comparison and verification: - Nasal Respiratory Airway Analysis Test: VISIONAIR™ AI application confirmation of successful segmentation of the Internal Nasal Valve and Nasal Cavum, and image manipulation/loading functions.
- CT vs Segmentation Accuracy Test: Comparison of endoscopic image cross-sectional areas segmented by VISIONAIR™ vs. cross-sectional areas of the same anatomical regions marked on CT scans (details of comparison not provided).
- VISIONAIR™ AI Segmentation Accuracy Test: Comparison of segmented endoscopic images by VISIONAIR™ vs. segmented endoscopic images by experienced clinicians (details of comparison not provided). |
| User Validation | User Validation Test: Validation of the entire VISIONAIR™ system by clinicians, including successful verification of all accessible features. |
| Substantial Equivalence | Concluded to be substantially equivalent to the predicate device in indication for use, performance, technology, features, principles of operation, and components. |
Detailed Study Information (Based on available text):
-
Sample Size used for the test set and the data provenance:
- Test Set Sample Size: Not specified for any of the listed tests.
- Data Provenance: Not specified (e.g., country of origin). The document mentions "endoscopic images" and "CT scans" were used, but no details on their origin or whether they were retrospective/prospective.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g., radiologist with 10 years of experience):
- Number of Experts: Not specified.
- Qualifications of Experts: The "VISIONAIR™ AI Segmentation Accuracy Test" mentions "experienced clinicians" were used for comparison, but their specific qualifications (e.g., specialty, years of experience) are not provided.
-
Adjudication method (e.g. 2+1, 3+1, none) for the test set:
- Not specified. The document only mentions "comparison" in the segmentation accuracy tests.
-
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:
- MRMC Study: Not explicitly described as an MRMC comparative effectiveness study where human readers' performance with and without AI assistance is measured.
- Effect Size: Not provided. The document mentions "User Validation Test" by clinicians and "comparison of segmented endoscopic images by VISIONAIR™ application vs. segmented endoscopic images by experienced clinicians," but not an assessment of human reader improvement.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- The "VISIONAIR™ AI application confirmation of successful segmentation" in the Nasal Respiratory Airway Analysis Test, and the "CT vs Segmentation Accuracy Test" and "VISIONAIR™ AI Segmentation Accuracy Test" imply a standalone evaluation of the algorithm's performance in segmentation against various ground truths (CT scans, experienced clinicians' segmentations). However, exact methodology and metrics are not detailed.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- CT Scans: Used as a ground truth for cross-sectional areas in the "CT vs Segmentation Accuracy Test." This implies measurements from CT scans were considered the reference.
- Experienced Clinician Segmentations: Used as a ground truth for segmented endoscopic images in the "VISIONAIR™ AI Segmentation Accuracy Test." This suggests individual or consensus segmentations by clinicians served as the reference.
- Implicitly, other tests depend on functional specifications and user observation.
-
The sample size for the training set:
- Not specified in the document. The document refers to the "trained AI algorithm" but does not give details about its training data.
-
How the ground truth for the training set was established:
- Not specified in the document.
§ 868.1800 Rhinoanemometer.
(a)
Identification. A rhinoanemometer is a device used to quantify the amount of nasal congestion by measuring the airflow through, and differential pressure across, a patient's nasal passages.(b)
Classification. Class II (performance standards).