(226 days)
EyeArt is indicated for use by healthcare providers to automatically detect more than mild diabetic retinopathy and visionthreatening diabetic retinopathy (severe non-proliferative diabetic retinopathy or proliferative diabetic retinopathy and/or diabetic macular edema) in eyes of adults diabetes who have not been previously diagnosed with diabetic retinopathy. EyeArt is indicated for use with Canon CR-2 Plus AF, and Topcon NW400 caneras.
EyeArt is a software as a medical device that consists of three components - Client, Server, and Analysis Computation Engine. A retinal fundus camera, used to capture retinal fundus images of the patient, is connected to a computer where the EyeArt Client software is installed. The EyeArt Client software provides a graphical user interface (GUI) that allows the EyeArt operator to transfer the appropriate fundus images to and receive results from the remote EyeArt Analysis Computation Engine through the EyeArt Server. The EyeArt Analysis Computation Engine is installed on remote computer(s) in a secure data center and uses artificial intelligence algorithms to analyze the fundus images and return results. EyeArt is intended to be used with retinal fundus images of resolution 1.69 megapixels or higher captured using one of the indicated retinal fundus cameras (Canon CR-2 AF, Canon CR-2 Plus AF, and Topcon NW400) with 45 degrees field of view. EyeArt is specified for use with two retinal fundus images per eye: optic nerve head (ONH) centered and macula centered. For each patient eye, the EyeArt results separately indicate whether "more than mild diabetic retinopathy (mtmDR)" and "vision-threatening diabetic retinopathy (vtDR)" are detected.
Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:
Acceptance Criteria and Device Performance
The provided document does not explicitly list pre-defined acceptance criteria in a separate table with specific numerical thresholds for sensitivity, specificity, etc. However, the performance metrics reported in the tables and the concluding statement ("The results of this prospective study support a determination of substantial equivalence between EyeArt v2.2.0 and EyeArt v2.1.0 and support the addition of the Topcon NW400 camera to the IFU statement" and "The results of this retrospective study support a determination of substantial equivalence between EyeArt v2.2.0 and EyeArt v2.1.0") imply that the demonstrated performance met the FDA's expectations for substantial equivalence to the predicate device.
The reported device performance for EyeArt v2.2.0 with the new Topcon NW400 camera and previous Canon cameras for detecting "more than mild diabetic retinopathy (mtmDR)" and "vision-threatening diabetic retinopathy (vtDR)" is summarized below. It's important to note that the comparison is implicit against the performance of the predicate device (EyeArt v2.1.0) and general expectations for diagnostic devices.
Table of Reported Device Performance (Prospective Study EN-01b):
| Metric | mtmDR (Canon CR-2 AF/Plus AF) | mtmDR (Topcon NW400) | vtDR (Canon CR-2 AF/Plus AF) | vtDR (Topcon NW400) |
|---|---|---|---|---|
| Sensitivity | 95.9% [90.4% - 100%] (70/73) | 94.4% [88.3% - 98.8%] (68/72) | 96.8% [90.0% - 100%] (30/31) | 96.8% [89.5% - 100%] (30/31) |
| Specificity | 86.4% [81.2% - 91.1%] (216/250) | 91.1% [86.8% - 94.8%] (226/248) | 91.7% [87.7% - 95.2%] (266/290) | 91.6% [87.5% - 95.1%] (263/287) |
| PPV | 67.3% [55.9% - 77.4%] (70/104) | 75.6% [64.6% - 85.4%] (68/90) | 55.6% [39.2% - 72.0%] (30/54) | 55.6% [38.0% - 72.1%] (30/54) |
| NPV | 98.6% [96.9% - 100%] (216/219) | 98.3% [96.4% - 99.6%] (226/230) | 99.6% [98.8% - 100%] (266/267) | 99.6% [98.5% - 100%] (263/264) |
| Best-case Sens. (mtmDR) | 95.9% [90.5% - 100.0%] (71/74) | 94.6% [88.6% - 98.8%] (70/74) | 96.8% [90.9% - 100.0%] (30/31) | 96.8% [89.5% - 100.0%] (30/31) |
| Worst-case Sens. (mtmDR) | 94.6% [88.9% - 98.8%] (70/74) | 91.9% [84.8% - 97.3%] (68/74) | 96.8% [90.0% - 100.0%] (30/31) | 96.8% [89.5% - 100.0%] (30/31) |
| Best-case Spec. (mtmDR) | 86.5% [81.3% - 91.2%] (218/252) | 91.3% [87.1% - 94.9%] (230/252) | 91.8% [87.9% - 95.3%] (269/293) | 91.8% [87.8% - 95.1%] (269/293) |
| Worst-case Spec. (mtmDR) | 85.7% [80.6% - 90.3%] (216/252) | 89.7% [85.05% - 93.3%] (226/252) | 90.8% [86.7% - 94.7%] (266/293) | 89.8% [85.7% - 93.6%] (263/293) |
Study Information:
1. Sample Sizes and Data Provenance:
-
Test Set (Prospective Study EN-01b):
- Accuracy Analysis: 336 eyes from 171 participants.
- Precision Analysis: 264 eyes from 132 participants.
- Data Provenance: Prospective, multi-center clinical study conducted in the United States (implied by FDA submission and context).
-
Test Set (Retrospective Study EN-01):
- Accuracy Analysis: 1310 eyes from 655 participants.
- Data Provenance: Retrospective, utilizing data already collected from the EyeArt pivotal multi-center clinical study (Protocol EN-01). Geographic origin not explicitly stated but likely United States due to the FDA context.
2. Number of Experts and Qualifications for Ground Truth:
- Number of Experts: Not explicitly stated as a count of individual experts, but the reference standard was determined by "experienced and certified graders" at the University of Wisconsin Reading Center (WRC).
- Qualifications of Experts: "experienced and certified graders" at the University of Wisconsin Reading Center (WRC).
3. Adjudication Method for the Test Set:
- The text states the ground truth was determined by "experienced and certified graders at the University of Wisconsin Reading Center (WRC)." It does not specify a numerical adjudication method (e.g., 2+1, 3+1). It implies a consensus or designated "grader" process, but the specifics of how disagreements (if any) were resolved are not detailed.
4. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:
- No, an MRMC comparative effectiveness study comparing human readers with AI vs. without AI assistance was not reported. The studies focused on the standalone diagnostic accuracy of the EyeArt device against a clinical reference standard (WRC grading).
5. Standalone Performance (Algorithm Only):
- Yes, standalone performance was done. The reported sensitivity, specificity, PPV, and NPV are measures of the EyeArt algorithm's performance in automatically detecting DR without direct human interpretation of the images in the diagnostic loop. The "EyeArt operator" described in the device description assists in image acquisition and receiving results, but the detection itself is algorithmic.
6. Type of Ground Truth Used:
- Expert Consensus / Clinical Reference Standard: The ground truth was established by a "Clinical reference standard (CRS)" which was determined by "experienced and certified graders at the University of Wisconsin Reading Center (WRC)" based on dilated 4-widefield stereo fundus images per the Early Treatment for Diabetic Retinopathy Study (ETDRS) severity scale. This is a form of expert consensus derived from a clinical gold standard imaging modality.
7. Sample Size for the Training Set:
- Not specified in the provided text. The document states that the six sites for the prospective study "did not contribute data used for training or development of EyeArt," implying a separate training dataset was used, but its size is not disclosed.
8. How Ground Truth for Training Set was Established:
- Not explicitly stated in the provided text. Given the consistency in methodology for the test set, it is highly probable that the ground truth for the training set was established similarly, using expert grading by ophthalmology specialists or a reading center. However, this is an inference, not directly stated for the training set.
§ 886.1100 Retinal diagnostic software device.
(a)
Identification. A retinal diagnostic software device is a prescription software device that incorporates an adaptive algorithm to evaluate ophthalmic images for diagnostic screening to identify retinal diseases or conditions.(b)
Classification. Class II (special controls). The special controls for this device are:(1) Software verification and validation documentation, based on a comprehensive hazard analysis, must fulfill the following:
(i) Software documentation must provide a full characterization of technical parameters of the software, including algorithm(s).
(ii) Software documentation must describe the expected impact of applicable image acquisition hardware characteristics on performance and associated minimum specifications.
(iii) Software documentation must include a cybersecurity vulnerability and management process to assure software functionality.
(iv) Software documentation must include mitigation measures to manage failure of any subsystem components with respect to incorrect patient reports and operator failures.
(2) Clinical performance data supporting the indications for use must be provided, including the following:
(i) Clinical performance testing must evaluate sensitivity, specificity, positive predictive value, and negative predictive value for each endpoint reported for the indicated disease or condition across the range of available device outcomes.
(ii) Clinical performance testing must evaluate performance under anticipated conditions of use.
(iii) Statistical methods must include the following:
(A) Where multiple samples from the same patient are used, statistical analysis must not assume statistical independence without adequate justification.
(B) Statistical analysis must provide confidence intervals for each performance metric.
(iv) Clinical data must evaluate the variability in output performance due to both the user and the image acquisition device used.
(3) A training program with instructions on how to acquire and process quality images must be provided.
(4) Human factors validation testing that evaluates the effect of the training program on user performance must be provided.
(5) A protocol must be developed that describes the level of change in device technical specifications that could significantly affect the safety or effectiveness of the device.
(6) Labeling must include:
(i) Instructions for use, including a description of how to obtain quality images and how device performance is affected by user interaction and user training;
(ii) The type of imaging data used, what the device outputs to the user, and whether the output is qualitative or quantitative;
(iii) Warnings regarding image acquisition factors that affect image quality;
(iv) Warnings regarding interpretation of the provided outcomes, including:
(A) A warning that the device is not to be used to screen for the presence of diseases or conditions beyond its indicated uses;
(B) A warning that the device provides a screening diagnosis only and that it is critical that the patient be advised to receive followup care; and
(C) A warning that the device does not treat the screened disease;
(v) A summary of the clinical performance of the device for each output, with confidence intervals; and
(vi) A summary of the clinical performance testing conducted with the device, including a description of the patient population and clinical environment under which it was evaluated.