Search Results

CASIA2 is a non-contact, high-resolution tomographic and biomicroscopic device intended for the in vivo imaging and measurement of ocular structures in the anterior segment. CASIA2 measures corneal thickness, anterior chamber depth and lens thickness.

The Tomey Cornea/Anterior Segment OCT CASIA2 (CASIA2) is a non-contact, high resolution tomographic and biomicroscopic device indicated for in vivo imaging of ocular structures in the anterior segment. The Tomey Cornea/Anterior Segment OCT CASIA2 is indicated as an aid in the visualization and measurement of anterior segment findings. CASIA2 measures corneal thickness, anterior chamber depth and lens thickness.

This medical device product has functions subject to FDA premarket review (corneal thickness, curvature, anterior chamber depth and lens thickness) as well as functions that are not subject to FDA premarket review. For this application, for the (510(k) exempt functions that are not subject to FDA premarket review, FDA assessed those functions only to the extent that they either could adversely impact the safety and effectiveness of the overall device.

CASIA2 consists of several components: the main unit, AC input power source, a touch panel LCD monitor, an external hard drive (HDD), a mouse and a keyboard.

Here’s a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided FDA 510(k) clearance letter for the Tomey CASIA2.

Device: Tomey Cornea/Anterior Segment OCT (CASIA2)
Measurements evaluated: Central Corneal Thickness (CCT), Anterior Chamber Depth (ACD), and Lens Thickness (LT).

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state pre-defined quantitative acceptance criteria (e.g., "CCT agreement must be within X µm"). Instead, it focuses on demonstrating agreement and precision compared to a legally marketed reference device (LENSTAR LS900). The "acceptance criteria" can be inferred from the study's objective to show substantial equivalence through these performance metrics. The reported device performance is presented as the actual agreement (mean difference and 95% Limits of Agreement - LOA) and precision (Repeatability and Reproducibility %CV).

Therefore, the table below reflects the demonstrated performance and implicitly what was considered acceptable for substantial equivalence.

Note on "Acceptance Criteria": The document implies that meeting or exceeding the performance of the LS900 in terms of precision, and demonstrating good agreement via Bland-Altman analysis, constituted the acceptance for substantial equivalence.

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size for Test Set: A total of 224 subjects were enrolled and completed the study for precision and agreement testing.
  • 55 subjects in the normal group
  • 60 subjects in the cataract group
  • 109 subjects in the special eyes group (eyes without a natural lens or eyes containing artificial materials)
    N for specific analyses (e.g., Agreement Analysis for LT) varied based on acceptable scans (e.g., N=122 for LT agreement, N=138 for CCT/ACD precision, N=76 for LT precision).
• Data Provenance: The document does not explicitly state the country of origin. It indicates "The subjects of this study had no notable or unexpected/untoward assessments..." which suggests a single clinical site. However, no specific country is mentioned.
• Retrospective or Prospective: The study was a prospective clinical study, as subjects were "enrolled," "randomized," and "assigned" to configurations and sequences, and data was collected during the study.

3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications

• The document implies that the LENSTAR LS900 device itself served as the reference standard for establishing "ground truth" (or more accurately, the comparator for agreement) for the measured parameters.
• It states that "The clinical site had 3 device operators trained on the devices used in the study."
• Qualifications of Experts: The specific qualifications (e.g., radiologist, ophthalmologist, optometrist expertise, years of experience) of these 3 device operators are not explicitly stated in the provided text.

4. Adjudication Method for the Test Set

• The document states, "Additional scans were taken at the operator's discretion if image quality was unacceptable based on the device DFU and the Tomey CASIA2 Reference Guide and included, missing scans, truncated scans, image defocus, floaters, presence of eye blinks, eye motion, etc. Each device operator had up to 3 attempts to obtain an acceptable scan for each of the required scans."
• This suggests an operational approach to ensure data quality rather than a formal, independent adjudication process (e.g., 2+1/3+1 consensus by experts) for the measurements themselves. The "ground truth" was derived from the in-device measurements of the LS900, not a separate expert review. Therefore, there was no expert consensus-based adjudication method for the measurements.

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

• No, a traditional MRMC comparative effectiveness study was not done.
• The study design was focused on device-to-device agreement and precision (CASIA2 vs. LENSTAR LS900) rather than evaluating how human readers' performance (e.g., diagnostic accuracy) improved with or without AI assistance.
• The CASIA2 is described as a "tomographic and biomicroscopic device intended for the in vivo imaging and measurement of ocular structures," with software providing "quantitative outputs." It does not appear to be an AI-assisted diagnostic aid for image interpretation that would typically require an MRMC study to show human reader benefit.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance)

• The study primarily assessed the measurement performance of the CASIA2 device (algorithm/system) in generating quantitative outputs (CCT, ACD, LT) and compared these directly with a reference device. It's implied that these measurements are generated automatically by the device's software.
• The role of the human operators was to acquire an "acceptable scan" based on predefined image quality criteria, not to interpret the images or provide a human "answer" for comparison with an AI-generated reading.
• Therefore, the precision and agreement studies essentially represent the standalone performance of the CASIA2's measurement capabilities compared to the LS900.

7. Type of Ground Truth Used

• The "ground truth" (or clinical reference standard) for comparison was the measurements obtained from the legally marketed predicate/reference device, LENSTAR LS900.
• This is a device-based comparative ground truth, not expert consensus, pathology, or outcomes data. The study aimed to show that the CASIA2's measurements were interchangeable or highly agreeable with those from an established, cleared device.

8. Sample Size for the Training Set

• The document describes a clinical study for validation/testing of the updated software. It does not provide any information about a separate training set size for the development of the algorithms generating these quantitative measurements.
• It only mentions: "The device is a software upgraded version of the predicate K213265 that provides quantitative measurements. All quantitative measurements are derived from OCT images acquired with optical coherence tomography." This implies the software update incorporated algorithms to derive these measurements, but details on their development (including training data) are not provided in this 510(k) summary.

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

• As the document does not describe the training set or its development, there is no information provided on how the ground truth for any training set was established.

Ask a specific question about this device