Search Results

The iVue is a non-contact, high resolution tomographic imaging device. It is intended for in-vivo imaging, axial crosssectional, and three-dimensional imaging and measurement of anterior ocular structures, including retina, retinal nerve fiber layer, ganglion cell complex (GCC), optic disc, corneal epithelia, corneal stroma and anterior chamber of the eye. With the integrated normative database, the iVue is a quantitative tool for the comparison of retina, retinal nerve fiber layer, ganglion cell complex, and optic disc measurements to a database of known normal subjects. The iVue is indicated for use as a device to aid in the diagnosis, documentation, and management of ocular health and diseases in the adult population.

The iVue is used to capture, store, display and print spectral domain-optical coherence tomography (SD-OCT) images of the posterior and anterior structure of the eye. The device software includes a Normative Database (NDB), consisting of OCT data from a range of known normal subjects that can be used to compare a new patient's measurements in relation to the normal distribution.

iVue is a computer-controlled ophthalmic imaging system using either a laptop computer or "All-in-One" computer. For laptop systems there are two control box options of 120 or 230 volts. The control box interfaces between the motorized table column and the medical-grade power supply for the computer.

iVue System Key Functional Components:

• Scanner Head
• Computer
• Control Box
• Joystick and Chinrest Assembly
• Footswitch (optional)
• Motorized Table (optional)
• Cornea Adapter Module

The provided document is a 510(k) Premarket Notification for the Optovue iVue device, seeking clearance for a software modification that allows automated segmentation and measurement of corneal epithelial and stromal layer thickness. The document describes two clinical studies conducted to support the substantial equivalence claim.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:

1. Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state quantitative acceptance criteria in a table format with pass/fail thresholds. Instead, it presents the results of repeatability, reproducibility, and agreement studies, and then concludes whether these results demonstrate "good repeatability and reproducibility" and "good agreement."

Based on the summaries provided, the implicit acceptance criteria seem to be related to achieving clinically acceptable levels of repeatability, reproducibility, and agreement with manual measurements for corneal pachymetry, epithelial thickness, and stromal thickness, across various patient groups.

Table of Implicit Acceptance Criteria and Reported Device Performance:

2. Sample Sizes and Data Provenance

Study 1: Repeatability and Reproducibility (R&R) of Corneal Epithelial Thickness Mapping

• Sample Size (Test Set): 59 subjects (from 598 total acquired scans, 527 qualified for final analysis).
  • Normal Subjects: 12 subjects
  • Corneal Patients: 47 subjects (further stratified into 4 subgroups, each with 11-12 subjects: Contact Lens, Dry Eye, Post-Refractive Surgery, Keratoconus).
• Data Provenance: Not explicitly stated, but clinical studies are generally prospective. The multi-device, multi-operator protocol in Study 1 suggests it was a controlled, prospective collection. Given that Optovue is based in Fremont, CA, and the submission is to the FDA, it is highly probable the data was collected in the USA, although not explicitly stated.

Study 2: Agreement of Corneal Epithelial Thickness Mapping with iVue SD-OCT to Manual Measurement

• Sample Size (Test Set): 87 study eyes from 87 subjects.
  • Normal Subjects: 17
  • Contact Lens (normal): 16
  • Dry Eye: 18
  • Post-Laser Refractive Surgery: 20
  • Keratoconus: 16
• Data Provenance: Not explicitly stated, but the mention of "3 study sites" collecting OCT data suggests a prospective clinical study. Again, given the FDA submission, most likely USA.

3. Number of Experts and Qualifications for Ground Truth

• Study 1 (R&R): Ground truth was based on the iVue software's automated segmentation, with manual edits performed by operators when needed. The document states: "For qualified scans, the operators reviewed the thickness maps for obvious segmentation error and reviewed the individual corneal meridian images to verify segmentation for erroneous maps. Segmentation edit tools were used to perform manual correction and then the epithelial map was reprocessed. Noticeable segmentation errors were manually corrected by the operator and marked for "Manual Correction" in the case report form."

  • The qualifications of these "operators" are not explicitly detailed. It implies they are trained users of the device capable of identifying and correcting segmentation errors, but their specific certifications (e.g., ophthalmologist, optometrist, ophthalmic technician) or years of experience are not mentioned.
  • Number of operators: The study used "device-specific designated operator" for each of the three iVue devices, so a minimum of 3 operators.

• Study 2 (Agreement): Ground truth was established by manual measurements performed by "3 qualified graders."

  • The qualifications of these "qualified graders" are not specified (e.g., ophthalmologist, ophthalmic technician, years of experience). They used a "2-section caliper tool that was available in the previously cleared iVue software."

4. Adjudication Method for the Test Set

• Study 1 (R&R): There's no explicit mention of a formal multi-reader adjudication process (e.g., 2+1, 3+1). It states "manual edits were not required in Normal eyes scans, and ranged from 2.8% in Contact Lens sub-group to 17.9% in KCN sub-group...Noticeable segmentation errors were manually corrected by the operator and marked for "Manual Correction"". This suggests correction was done by a single operator per scan if needed, not a consensus or adjudicated process among multiple experts.
• Study 2 (Agreement): The "agreement to manual measurement of the software was evaluated based on the agreement between the software output and the manual measurements by 3 qualified graders." It appears each grader provided their measurement independently, and then the agreement between the software and their measurements (or a combined representation of their measurements, though "manual measurements" is singular in "agreement to manual measurement") was assessed. It doesn't describe an adjudication process amongst the 3 graders to establish a single ground truth for each case from the graders. The comparison is between the software and the individual manual measurements, then reported collectively.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• No MRMC comparative effectiveness study was done comparing human readers with AI vs. without AI assistance. The study focuses purely on the device's technical measurement performance (repeatability, reproducibility, and agreement with existing manual measurement methods). The iVue is a measurement device, not an AI diagnostic aid in the sense of 'improving human reader performance.'

6. Standalone (Algorithm Only) Performance

• Yes, standalone performance was evaluated. The core of both clinical studies is the performance of the iVue device's software (algorithm) in segmenting and measuring the corneal layers.
  • In Study 1, it assessed the repeatability and reproducibility of the "modified iVue software" directly.
  • In Study 2, it assessed the "agreement between the software output and the manual measurements." This is a direct evaluation of the algorithm's performance against a human-derived ground truth.

7. Type of Ground Truth Used

• Study 1 (R&R): The ground truth was the automated segmentation by the iVue software, potentially refined by "operator correction". This is a form of expert-corrected algorithmic output.
• Study 2 (Agreement): The ground truth was manual measurements performed by "qualified graders" using a digital caliper tool available in the predicate iVue software. This is a form of expert measurement. It's not pathology or outcomes data.

8. Sample Size for the Training Set

• Not specified in the provided document. The document describes clinical studies for validation and verification of a modified software, not the development or training of the original algorithm. Device approvals like 510(k) often focus on testing the finished product.

9. How the Ground Truth for the Training Set Was Established

• Not specified in the provided document. As the training set size is not mentioned, neither is the method for establishing its ground truth. This information would typically be found in the algorithm's development and internal validation documentation, not necessarily in a 510(k) summary focused on post-modification performance.

Ask a specific question about this device