Search Results

The Konan Specular Microscope XVII, CellChek 20, is a non-contact ophthalmic microscope, optical pachymeter, and camera intended for examination of the corneal endothelium and for measurement of the thickness of the cornea.

The Konan Specular Microscope XVII, CellChek 20, is a non-contact ophthalmic microscope, optical pachymeter, and camera intended for examination of corneal endothelium and for measurement of the thickness of the cornea. Cell counting and analysis program are included, and allow for analysis of the images of the cell distribution of the eye.

When photographing the corneal endothelium, the device performs the alignment and automatically focuses by capturing the reflected light from patient's eye with the camera. The device permits visual inspection and photography of the corneal endothelium and measurement of the corneal thickness without any object contacting the eye. It features focusing by means of infrared techniques, as well as computer-assisted cell counting and cell analysis capabilities. The computer functions are also used to aid in setting up the various features of the machine and to aid in photography. Photographic images are temporarily stored in the system's memory and can be preserved by using a printer.

The parts of the device that come into contact with a patient are the forehead rest and the chin rest. Their material is acrylonitrile butadiene styrene (ABS), the same material used in reference device and one with proven biocompatibility.

The function of the software installed in the device is to calculate mainly the cell density, the coefficient of variation of cell area and the percent hexagonality. In the manual methods, Actual identification if the cells and cell boundaries is done by the (physician) user. In the automatic method, the software detects the cells and cell boundaries, however, the user is given the opportunity to make corrections. In use, the user interacts with the software by visually placing dots in the center of cells as or by tracing cell boundaries as they appear or on a screen or uses the automatic algorithm.

The provided text describes the 510(k) summary for the Konan Specular Microscope XVII, CellChek 20. The bulk of the performance study provided is focused on demonstrating agreement, accuracy and precision with a reference device (NONCON ROBO PACHY F&A) and comparing various analysis methods within the subject device itself. It does not present acceptance criteria in a quantitative format nor directly compare the device performance against specific, pre-defined thresholds for acceptance. Instead, it focuses on comparative equivalence.

However, based on the provided clinical study data and the overall goal of demonstrating "substantial equivalence," we can infer the implicit "acceptance criteria" through the statistical results presented, particularly the correlations and comparisons against the reference device and internal method consistency. The study design is primarily a comparative study rather than a standalone AI performance evaluation or a multi-reader, multi-case study with human readers.

Here's an attempt to structure the information based on the request, interpreting the "acceptance criteria" from the study's conclusions on "strong positive correlation" and "superiority/equivalence in precision."

Implicit Acceptance Criteria and Reported Device Performance

The acceptance criteria are implicitly based on demonstrating "strong positive correlation" and "superiority to or equivalence with" the predicate/reference device and consistency across different analysis methods within the subject device for key metrics.

Table of Implicit Acceptance Criteria and Reported Device Performance

• CD: R² = 0.9193 (Strong Positive Correlation)
• CV: R² = 0.5677 (Strong Positive Correlation, although the R² value quantitatively is moderate)
• HEX: R² = 0.2460 (Strong Positive Correlation, although the R² quantitatively is weak)

Peripheral Area:

• CD: R² = 0.9209 (Strong Positive Correlation)
• CV: R² = 0.4974 (Strong Positive Correlation, although the R² quantitatively is moderate)
• HEX: R² = 0.1439 (Weak Positive Correlation, as stated in Table 14 conclusion) |
  | Agreement Across Internal Methods | Strong positive correlation (high R²) and acceptable limits of agreement between different analysis methods (e.g., Trace, Auto Trace) and the Center Method within the subject device. | Center Area - Center Method vs. Others:
• Trace: CD, CV, HEX all showed "Strong Positive Correlation" (R² range 0.8097 to 0.9973)
• Auto Trace: CD, CV showed "Strong Positive Correlation" (R²=0.8479, 0.5682), HEX showed "Positive Correlation" (R²=0.4050)
• Auto Center: CD, CV, HEX all showed "Strong Positive Correlation" (R² range 0.6055 to 0.8504)
• Flex Center: CD, CV, HEX all showed "Strong Positive Correlation" (R² range 0.7587 to 0.9971)
• Auto Flex Center: CD, CV showed "Strong Positive Correlation" (R²=0.8480, 0.6499), HEX showed "Positive Correlation" (R²=0.4564). This aligns with the "Strong/Positive Correlation" statements in Table 15. |
  | Precision (Repeatability & Reproducibility) vs. Reference Device | Repeatability and Reproducibility of the subject device should be superior or equivalent to the predicate/reference device. | Center Area (Center Method vs F&A): CD, CV, HEX for CellChek 20 (Subject Device) were "superior to" F&A (Reference Device) based on lower Repeatability SD, Reproducibility SD, and narrower limits. This general superiority was maintained across other CellChek 20 analysis methods (Auto Trace, Auto Center, Auto Flex Center are "very superior," and Trace/Flex Center are "partially superior or equivalent"). |

Detailed Study Information:

1. Sample size used for the test set and the data provenance:

   • Sample Size: 80 subjects (patients).
   • Data Provenance: The study was a "prospective clinical study" conducted at an "ophthalmic clinic" that agreed to cooperate. The country of origin is not explicitly stated in the provided text, but Konan Medical, Inc. is based in Japan. Without that specific detail in the document, it's assumed to be from a clinical setting, but the geographical location of that clinic isn't mentioned beyond "ophthalmic clinic."

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

   • The study states: "Three analysts analyzed the examinees' images with each of the 6 methods, Center Method, Auto Center Method, Trace Method, Auto Trace Method, Flex Center Method, Auto Flex Center Method, of CellChek 20, and Center Method of F&A."
   • The qualifications of these "analysts" are not specified. It is implied they are trained individuals capable of performing these analyses, but their explicit qualifications (e.g., "radiologist with 10 years of experience" or similar) are not provided. The document mentions that in manual methods, "Actual identification if the cells and cell boundaries is done by the (physician) user," implying physicians are involved in the process, but it doesn't state if the "three analysts" were physicians.

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

   • The document does not describe any adjudication method for establishing a single "ground truth" or reference standard. The "three analysts" each performed their analysis, and the study then compares the subject device's methods (analyzed by these three) with the reference device's method (also analyzed by these three). It appears to be a direct comparison of device outputs without an independent, adjudicated ground truth.

4. 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, a multi-reader, multi-case (MRMC) comparative effectiveness study involving human readers improving with AI assistance vs. without AI assistance was not performed. The study evaluates the device's measurements (Corneal Endothelial Cell Density, Coefficient of Variation, and Hexagonality) against a reference device and compares different analysis methods within the device. There is no mention of human-in-the-loop performance improvement.

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

   • Yes, in a sense. The study assesses the performance of the device's various analysis methods (Trace, Auto Trace, Center, Auto Center, Flex Center, Auto Flex Center) in terms of agreement, accuracy, and precision compared to a reference device. While analysts operate the device and its software, the measurements themselves are output by the device's algorithms. The "Auto Trace," "Auto Center," and "Auto Flex Center" methods specifically involve automated detection, and their performance is evaluated. Thus, the performance of the built-in algorithms for measurement is evaluated.

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

   • The "ground truth" or reference standard for comparison was the output from a legally marketed predicate/reference device, the NONCON ROBO PACHY F&A (K062763), specifically using its Center Method, which was also analyzed by the "three analysts." It is an instrument-based reference standard rather than an independent clinical ground truth like pathology or long-term patient outcomes.

7. The sample size for the training set:

   • The document describes a clinical performance study for the Konan Specular Microscope XVII, CellChek 20 (test set of 80 subjects). It does not provide any information regarding the training set size or methodology for the software algorithms embedded in the device. The algorithms were likely developed and validated internally by the manufacturer prior to this submission study.

8. How the ground truth for the training set was established:

   • As no information about the training set is provided, how its ground truth was established is also not described in this document.

Ask a specific question about this device