Search Results

The Spectralis HRA+OCT is a non-contact ophthalmic diagnostic imaging device. It is intended for viewing the posterior segment of the eye, including two- and three-dimensional imaging, cross-sectional imaging, fundus photography, and fluorescence imaging (fluorescein, indocyanine green and autofluorescence), and to perform measurements of ocular anatomy and ocular lesions. The device is indicated as an aid in the detection and management of various ocular diseases including: age-related macular degeneration, macular edema, diabetic retinopathy, retinal and choroidal vascular diseases, glaucoma, and for viewing geographic atrophy as well as changes in the eye that result from neurodegenerative diseases. The Spectralis HRA+OCT includes a retinal nerve fiber layer thickness normative database, which is used to quantitatively compare the retinal nerve fiber layer in the human retina to a database of Caucasian normal subjects; the classification result is valid only for Caucasian subjects.

The Spectralis HRA+OCT is a real-time imaging system of the posterior segment of the human eye and for aiding in the assessment and management of various diseases of the posteriorsegment, such as age-related macular degeneration, diabetic retinopathy, and glaucoma. The device is a combination of optical coherence tomography (OCT) with confocal scanning laser ophthalmoscopy (cSLO). OCT imaging includes high-resolution cross-sectional imaging of ocular structures (e.g., retina, macula, optic nerve head); cSLO imaging includes high-resolution and dynamic infrared reflectance, blue reflectance, fluorescein angiography, indocvanine green angiography, and autofluorescence imaging. OCT images and cSLO images are acquired simultaneously and are viewed side-by-side on the computer screen. Images are acquired and stored using Spectralis operation software, which runs on a standard personal computer. Spectralis components include a laser scanning camera mount with headrest, operation panel, power supply box, operation software, and host computer. A MultiColor option has been added to provide additional green reflectance imaging and a "composite color" image, which provides a different view of the features of the eye. This composite color image is not the same as fundus color photo.

The provided 510(k) summary for the Spectralis HRA+OCT device primarily focuses on demonstrating substantial equivalence to a predicate device and safety/performance through bench testing and a normative database study. It does not contain a typical acceptance criteria table with corresponding reported performance for a specific algorithm or AI model, nor does it describe a study specifically designed to prove a device meets such criteria in the context of AI performance.

Instead, the clinical evaluation section describes studies related to the device's measurement accuracy, repeatability, reproducibility, and the establishment of a normative database for Retinal Nerve Fiber Layer (RNFL) thickness, which is then used for classification.

Here's an attempt to extract and present the information based on the provided text, acknowledging the limitations in scope for AI-specific acceptance criteria and studies:

1. Table of "Acceptance Criteria" and Reported Device Performance

As this is a 510(k) for an imaging device with a normative database feature, the "acceptance criteria" are not framed in terms of AI performance metrics (like sensitivity, specificity, AUC). Instead, the performance is demonstrated through the characteristics of the normative database and the agreement with a predicate device.

• Age adjustment: Linear regression of RNFL thickness vs. age was performed for various sectors; negative slopes showed decrease with age and were adjusted; insignificant positive slopes were not adjusted. |
  | Agreement with Predicate Device (Stratus) | A good linear correlation between RNFL thickness measurements with Spectralis and Stratus devices was found for healthy and glaucoma subjects across all measurement regions. Slopes and intercepts of regression lines were in the neighborhood of 1 and 0, respectively, though with some variation (indicating they should not be used interchangeably, which is noted as in agreement with published literature). |
  | Performance in Disease (Qualitative) | Case Reports and Case Series of eyes with various pathologies showed no artifacts, no unexpected RNFL thickness measurement results, and no unexpected classification results. RNFL thickness was found to be predictably decreased in glaucoma subjects. (This is a qualitative statement of expected behavior rather than a strict quantitative criterion). |
  | Safety and Electrical Compliance | Tested according to IEC 60601-1, IEC 60601-1-2, IEC 60601-1-4, and IEC 60825-1, meeting all requirements. Classified as a Class 1 laser product per 21 CFR §1040.10. |

2. Sample size used for the test set and the data provenance

• Reproducibility and Repeatability: "human volunteers" - specific number not provided.
• Normative Database (Acts as a reference/test for classification): 201 subjects of Caucasian origin.
  • Provenance: Enrolled in a patient registry (implied prospective data collection for the purpose of the database). All subjects were described as "normal" based on specific criteria.
  • Limitations noted: Sample size (201), particularly small for extreme age groups (70 years), Caucasian ethnicity only, and inclusion of refractive errors from +5 to -7 diopters.
• Agreement Study with Predicate:
  • Healthy subjects: n=101
  • Glaucoma patients: n=183
  • Provenance: Not explicitly stated as retrospective or prospective, but implies direct examination for the study.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• Normative Database: "Screening for entry into the study included patient history and physical examination to determine if eyes were 'normal' by two ophthalmologists." No specific years of experience are provided, but "ophthalmologist" implies a qualified medical doctor specializing in ophthalmology.
• Agreement Study with Predicate: No explicit mention of experts establishing ground truth; the study compared measurements between two devices. The classification of "glaucoma patients" would imply diagnosis by medical professionals.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Normative Database: For determining "normality," two ophthalmologists made the determination. It's unclear if there was an adjudication process if their opinions differed ("2+0" if they agreed, no explicit mention of dispute resolution).
• Other studies: Adjudication methods are not mentioned.

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

• No MRMC study of human readers with vs. without AI assistance was reported. This device predates the widespread use of AI in medical imaging interpretation as we know it today. The "normative database" itself is a classification tool provided by the device, not an AI interpreting images for a human. It provides a color-coded classification based on measured RNFL thickness compared to the database.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Yes, in spirit, the normative database is a standalone "algorithm" for classification. The device automatically measures RNFL thickness and then classifies it against the stored normative database (e.g., as within or outside the 1st or 5th percentile). The performance of this classification is implicitly demonstrated by the database's construction, reproducibility of measurements, and the statement that RNFL thickness was "predictably decreased in glaucoma subjects." However, it is not presented with traditional standalone diagnostic performance metrics (e.g., sensitivity, specificity, AUC).

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

• Normative Database: Ground truth for "normality" was established by expert consensus/clinical examination by two ophthalmologists.
• Agreement Study with Predicate: The "ground truth" for comparison was the measurements from an established predicate device (Stratus OCT). For classifying subjects as "healthy" or "glaucoma," the ground truth would have been clinical diagnosis.

8. The sample size for the training set

• The document describes the normative database as being built from 201 subjects. This serves as the reference data against which subsequent patient measurements are compared for classification. In a machine learning context, this database acts effectively as the "training" or "reference" set for the classification rules (i.e., percentiles for normality).

9. How the ground truth for the training set was established

• For the normative database (which serves as the "training set" for the classification logic), the ground truth for "normality" was established by two ophthalmologists. Subjects were included if they had "no history of glaucoma, normal intraocular pressure, normal visual field, normal appearance of optic disc, etc." based on patient history and physical examination.

Ask a specific question about this device