(86 days)
The P200TE is a non-contact scanning laser ophthalmoscope and optical coherence tomographer intended for in-vivo viewing and digital imaging of posterior ocular structures, including the retinal nerve fiber layer and optic disc. It is indicated for producing high-resolution, widefield, en face reflectance images, autofluorescence images, and axial, cross-sectional images of the posterior ocular structures.
The P200TE is based on Scanning Laser Ophthalmoscope (SLO) technology which scans in two dimensions over the retina. Light reflected from the retina is detected and transformed into a digital image. Images may be stored and subsequently reviewed. The P200TE allows images to be captured in the following imaging modes: Reflectance imaging, Autofluorescence imaging, Optical coherence tomography imaging. The P200TE instrument uses red and green laser illumination for reflectance imaging, enabling it to image pathology throughout the layers of the retina, from the sensory retina and nerve fiber layer, through the retinal pigment epithelium (RPE) and down to the choroid. The image can be separated to present the distinct retinal sub-structures associated with the individual imaging wavelengths. The P200TE instrument uses green laser illumination to excite autofluorescence (AF) emission from the naturally occurring lipofuscin in the human fundus. The P200TE instrument uses infrared laser illumination for reflectance imaging simultaneously with OCT imaging. Infra-red reflectance images are used to track eye position during OCT imaging and are not available to the user. The P200TE instrument uses infrared superluminescent diode (SLD) illumination for optical coherence tomography allowing a depth profile of the reflectance of the human fundus to be recorded. The P200TE images the eye via two ellipsoidal mirrors arranged so that a focal point of one of the mirrors coincides with a focal point of the other mirror; a mirrored scanner is also located at this common focal point. The pupil of the subject's eye is placed at one of the other focal points. A second mirrored scanner is located at the remaining focal point; a laser or SLD reflected off this scanner is relayed onto the second scanner by the first ellipsoidal mirror and from there is reflected through the pupil and into the eye by the second ellipsoidal mirror. The second scanning element is different for OCT and SLO imaging. The energy reflected back from the retina, or emitted by fluorophores returns through the same path to the detectors; the images are generated from the captured detector data. This is operationally installed to be networked with computer peripherals and proprietary software that facilitate the storage, management and viewing of the retinal images. The images are captured by the scan head under operator control and then automatically saved to the image server that uses a database structure to hold the images and patient information. For subsequent image review, a number of viewing PC's are connected via a local area network to the image server. The patient records and images are then accessible in a distributed format suited to the physical layout of the eye-care practice.
The provided text focuses on establishing substantial equivalence to predicate devices rather than defining specific acceptance criteria for performance metrics. However, it does describe a clinical study conducted to demonstrate the device's similarity to its predicate for qualitative clinical use.
Here's an analysis based on the information provided:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly state numerical acceptance criteria for device performance. Instead, the "acceptance criteria" for the study were implicitly tied to demonstrating non-inferiority and similarity in image quality for qualitative clinical use compared to the predicate device.
| Acceptance Criteria (Implied) | Reported Device Performance |
|---|---|
| Image quality for qualitative clinical use is similar between P200TE and Spectral OCT/SLO. | - Wilcoxon signed-rank test analyses showed non-inferiority for image quality between P200TE and Spectral OCT/SLO. |
- Linearly weighted Cohen's kappa analyses showed adequate inter-grader agreement in scoring.
- Results "showed that the image quality for purposes of qualitative clinical use are similar between the P200TE and the Spectral OCT/SLO." |
| Compliance with recognized consensus standards (Non-Clinical) | - P200TE complied with ISO 15004-2:2007 (Light hazard protection), IEC60825-1 (Safety of Laser Products), and IEC 62366 (Usability engineering). - Results of bench testing demonstrated compliance with relevant recognized consensus standards. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: 35 participants (32 study eyes).
- Data Provenance: Prospective, observational study at a single U.S. clinical site.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- Number of Experts: Three independent, masked graders.
- Qualifications of Experts: The document does not specify the exact qualifications (e.g., "radiologist with 10 years of experience"). However, they were referred to as "graders" at a "third-party reading center" and "qualitatively evaluated the images based on pre-specified grading criteria... for qualitative clinical use," suggesting they were likely ophthalmologists, optometrists, or trained image graders with expertise in ocular imaging.
4. Adjudication Method for the Test Set
- The document states that "three independent, masked graders qualitatively evaluated the images." It then notes that "The grading results from the two devices were then compared using one-tailed Wilcoxon signed-rank test analyses... and linearly weighted Cohen's kappa analyses."
- The use of "independent" graders and the subsequent statistical comparison (including inter-grader agreement with kappa) suggests a consensus or comparison approach, rather than a strict 2+1 or 3+1 adjudication rule where disagreements are explicitly resolved by a third party. The primary goal was to assess agreement and similarity across graders and between devices.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- Yes, a form of MRMC study was conducted. Three independent graders evaluated images from both the P200TE and the predicate device (Spectral OCT/SLO) across multiple cases (32 study eyes from 35 participants).
- Effect Size of Human Readers' Improvement with AI vs. without AI Assistance: This information is not applicable/not provided. The study described is a comparison of two imaging devices (P200TE vs. Spectral OCT/SLO) in terms of image quality for qualitative clinical use, not an AI-assisted interpretation study. The P200TE is an imaging device, not an AI algorithm designed to improve human reading.
6. Standalone Performance (Algorithm Only Without Human-in-the-Loop Performance)
- Not applicable/not provided. The P200TE is an ophthalmoscope and optical coherence tomographer, a diagnostic imaging device. Its performance is assessed in terms of image quality for human interpretation, not as a standalone algorithm performing automated detection or diagnosis.
7. Type of Ground Truth Used
- The ground truth for the qualitative evaluation of image quality was established by expert assessment/consensus based on "pre-specified grading criteria... based on the presence or absence of clinically relevant structures and anatomic boundaries and on the overall image clarity necessary for qualitative clinical use."
- The "truth" for the study was not about disease presence (e.g., pathology, outcomes data) but rather the quality of the images produced by the device as judged by experts.
8. Sample Size for the Training Set
- Not provided/not applicable. The P200TE is an imaging device (hardware and embedded software) for image acquisition, not an AI model that requires a training set of images for learning. The document describes clinical validation of the device's output, not the development of a trained algorithm.
9. How the Ground Truth for the Training Set Was Established
- Not provided/not applicable. As mentioned above, this device is not an AI model that underwent a training phase on a labeled dataset. The study focused on validating the image quality of the P200TE compared to its predicate for human interpretation.
§ 886.1570 Ophthalmoscope.
(a)
Identification. An ophthalmoscope is an AC-powered or battery-powered device containing illumination and viewing optics intended to examine the media (cornea, aqueous, lens, and vitreous) and the retina of the eye.(b)
Classification. Class II (special controls). The device, when it is an AC-powered opthalmoscope, a battery-powered opthalmoscope, or a hand-held ophthalmoscope replacement battery, is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 886.9.