(226 days)
The IRIS Intelligent Retinal Imaging Systems is a comprehensive web-based software system application intended for use in storing, managing, and displaying patient data, diagnostic data and images from computerized diagnostic instruments or systems. Original and color amplified images can be viewed by trained healthcare professionals.
The IRIS software is a software as a service application that is hosted on the internet which allows clinicians the ability to scan a patient's retina with a fundus camera, transmit the images up to a website and offer an opinion on the scans. It also allows users the ability to input data relative to a visual fields exam. The combination of a fundus picture and visual fields data allows a licensed/credentialed clinician to evaluate the patient for glaucoma.
The provided document is a 510(k) summary for the IRIS Intelligent Retinal Imaging System. This summary states that no performance data was required or provided for the device to support substantial equivalence. The submission relies solely on software validation and verification to demonstrate that the software performs as intended.
Therefore, many of the specific questions about acceptance criteria, study details, sample sizes, expert involvement, and ground truth cannot be answered from the provided text.
Here's a breakdown of what can and cannot be answered:
1. A table of acceptance criteria and the reported device performance
- Cannot be answered from the text. The document explicitly states, "No performance data was required or provided."
2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Cannot be answered from the text. No test set performance data was provided or required.
3. 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)
- Cannot be answered from the text. No ground truth for a test set was established as no performance study was conducted.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
- Cannot be answered from the text. No test set performance data was provided or required.
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
- Cannot be answered from the text. No MRMC study was mentioned or performed. The device is described as an "Ophthalmic Image Management System" and "web-based software system application intended for use in storing, managing, and displaying patient data, diagnostic data and images." It does not appear to involve AI assistance for diagnosis, but rather provides tools for clinicians to review images and input their opinions.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Cannot be answered from the text. No standalone performance study was mentioned or performed.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
- Cannot be answered from the text. No ground truth was established for "performance data" as none was required or provided. The system allows licensed/credentialed clinicians to input their "discreet opinions" and "care plan based on grading levels chosen," implying human expert opinion is the primary diagnostic mechanism.
8. The sample size for the training set
- Cannot be answered from the text. The document describes the device as a software application for managing and displaying images and data, not an algorithm that requires a training set in the typical machine learning sense for diagnostic purposes.
9. How the ground truth for the training set was established
- Cannot be answered from the text. As above, the nature of the device does not suggest a training set for a diagnostic algorithm.
In summary: The IRIS Intelligent Retinal Imaging System was cleared based on substantial equivalence to predicate devices, focusing on its function as an image management and display system. The FDA did not require or receive performance data related to diagnostic accuracy from the device's algorithms (as it appears to be a viewing and management system, not an AI diagnostic tool itself), thus bypassing the need for a study with acceptance criteria, test sets, or ground truth establishment.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).