Search Results
Found 1 results
510(k) Data Aggregation
(274 days)
The Canon OCT-A1 is an optical coherence tomography system indicated for the in-vivo imaging and measurement of the retina, retinal nerve fiber layer, and optic disc as a tool and an aid in the diagnosis and management of retinal diseases by a clinician.
The Canon OCT-A1 *- is an Optical Coherence Tomography (OCT) system intended for use as a non-invasive imaging device for viewing and measuring ocular tissue structures with micrometer range resolution. The OCT-A1 is a computer controlled ophthalmic imaging system. The device scans the patient's eye using a low coherence interferometer to measure the reflectivity of retinal tissue. The cross sectional retinal tissue structure is composed of a sequence of A-scans. It has a traditional patient and instrument interface like most ophthalmic devices.
Here's a summary of the acceptance criteria and study details for the Canon OCT-A1, based on the provided document:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly state "acceptance criteria" with specific numerical targets in the context of device performance. Instead, it demonstrates substantial equivalence to a predicate device (Optovue RTVue XR OCT Avanti) by showing agreement and precision in measurements. The performance is reported in terms of mean differences and 95% Limits of Agreement (LOA) for agreement, and repeatability/reproducibility standard deviations (SD) and limits, along with Coefficient of Variation (CV%), for precision.
The key finding for agreement is that mean differences were small, often less than 10% of the thickness, with some wide LOAs particularly in diseased eyes. For precision, the Canon OCT-A1 showed comparable or better precision than the predicate device for many parameters, especially in normal and retinal disease groups, while occasionally showing more variability in glaucoma for some FRT and ONH parameters. The differences in FRT measurements (around 25 µm) were attributed to different segmentation methodologies (inclusion/exclusion of RPE).
Below is a summary table based on the provided data, reflecting the reported performance relative to the predicate rather than predefined absolute acceptance criteria.
| Measurement Parameter | Subject Population | Canon OCT-A1 Mean (SD) / Repeatability Limit (CV%) | Predicate Device Mean (SD) / Repeatability Limit (CV%) | Mean Difference (SD) / LOA (for Agreement) | Interpretive Summary of Performance against Predicate (for Precision where available) |
|---|---|---|---|---|---|
| Full Retinal Thickness (FRT) | |||||
| Central (µm) | Normal | 267.20 (19.396) / 17.37 (2.34%) | 248.89 (18.222) / 5.28 (0.76%) | 18.32 (5.056) / (8.20, 28.43) | Canon OCT-A1 had higher mean FRT (due to RPE inclusion) and generally wider LOA. For central FRT, predicate showed less variability (Repeatability Limit Ratio: 3.2893 in favor of Optovue). |
| Central (µm) | Glaucoma | 281.48 (35.758) / 18.17 (2.29%) | 261.59 (34.850) / 10.29 (1.41%) | 19.90 (24.654) / (-29.41, 69.21) | Similar trend as normal group for FRT. Predicate showed less variability (Repeatability Limit Ratio: 1.7658 in favor of Optovue). |
| Central (µm) | Retinal Disease | 301.31 (76.663) / 27.65 (3.30%) | 283.08 (87.687) / 32.29 (4.15%) | 18.23 (26.086) / (-33.94, 70.40) | Canon OCT-A1 showed slightly better repeatability (Repeatability Limit Ratio: 0.8565 in favor of Canon). |
| Other FRT parameters | All groups | (See Tables 1 & 5 for full details) | (See Tables 1 & 5 for full details) | Mean differences 10-30 µm, wide LOAs in some cases. | Mixed precision results, some parameters comparable, others one device better. |
| Retinal Nerve Fiber Layer (RNFL) Thickness | |||||
| TSNIT Average (µm) | Normal | 103.673 (9.5828) / 3.638 (1.247%) | 100.302 (7.9933) / 5.631 (2.001%) | 3.371 (4.9991) / (-6.628, 13.369) | Mean differences generally < 10%. Wide LOA for some sectors in diseased eyes. Canon OCT-A1 showed better precision (Repeatability Limit Ratio: 0.6460 in favor of Canon). |
| TSNIT Average (µm) | Glaucoma | 82.026 (16.5404) / 5.129 (2.275%) | 77.934 (15.3871) / 4.197 (1.967%) | 4.092 (7.4361) / (-10.781, 18.964) | Predicate device showed better precision (Repeatability Limit Ratio: 1.2223 in favor of Optovue). |
| TSNIT Average (µm) | Retinal Disease | 97.742 (14.7850) / 5.932 (2.145%) | 95.124 (13.4346) / 6.398 (2.398%) | 2.618 (4.6673) / (-6.717, 11.952) | Canon OCT-A1 showed better precision (Repeatability Limit Ratio: 0.9272 in favor of Canon). |
| Ganglion Cell Complex (GCC) Thickness | |||||
| Total (µm) | Normal | 90.934 (6.7238) / 1.740 (0.686%) | 94.548 (6.9738) / 2.773 (1.052%) | -3.614 (2.7775) / (-9.169, 1.941) | Canon OCT-A1 showed better precision (Repeatability Limit Ratio: 0.6275 in favor of Canon). |
| Total (µm) | Glaucoma | 82.138 (14.1996) / 6.901 (3.001%) | 85.419 (28.7960) / 18.896 (7.901%) | -5.486 (17.0916) / (-39.669, 28.697) | Canon OCT-A1 showed significantly better precision (Repeatability Limit Ratio: 0.3652 in favor of Canon). |
| Total (µm) | Retinal Disease | 92.972 (16.1076) / 4.917 (1.889%) | 98.009 (21.5411) / 8.220 (2.995%) | -5.671 (7.5186) / (-20.709, 9.366) | Canon OCT-A1 showed better precision (Repeatability Limit Ratio: 0.5982 in favor of Canon). |
| Optic Nerve Head (ONH) | |||||
| Disc Area (mm²) | Normal | 2.0482 (0.42161) / 0.1998 (3.4800%) | 2.0230 (0.38576) / 0.2942 (5.2436%) | 0.0252 (0.22361) / (-0.4220, 0.4725) | Mean differences for ONH parameters generally close to zero. Canon OCT-A1 showed better precision (Repeatability Limit Ratio: 0.6792 in favor of Canon). |
| Disc Area (mm²) | Glaucoma | 2.2282 (0.83476) / 0.9532 (15.6849%) | 2.0014 (0.38825) / 0.4238 (7.4635%) | 0.2268 (0.85919) / (-1.4916, 1.9452) | Predicate device showed better precision (Repeatability Limit Ratio: 2.2489 in favor of Optovue). |
| Disc Area (mm²) | Retinal Disease | 2.0591 (0.43518) / 0.2139 (3.6856%) | 1.9779 (0.37971) / 0.4350 (7.8835%) | 0.0812 (0.24191) / (-0.4026, 0.5650) | Canon OCT-A1 showed better precision (Repeatability Limit Ratio: 0.4917 in favor of Canon). |
| Other ONH parameters | All groups | (See Tables 4 & 8 for full details) | (See Tables 4 & 8 for full details) | Mean differences usually close to zero, some wide LOAs. | Mixed precision results, some parameters comparable, others one device better. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: 125 subjects, categorized into three groups:
- Normal Subjects: 37
- Glaucoma Subjects: 37
- Retinal Disease Subjects: 38
- Note: The number of evaluable eyes was "at least 99", with approximately 33 eligible subjects in each group. The tables explicitly state the N for each group as 37, 36 (for glaucoma RNFL, ONH) or 38, likely referring to the number of eyes rather than subjects for analysis.
- Data Provenance: Prospective, comparative, randomized, single-center study conducted in the US.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document does not mention the use of experts to establish a "ground truth" for the test set in the traditional sense (e.g., expert consensus on diagnosis or specific measurements for comparison against the device). This study is a comparative effectiveness study against a predicate device, focusing on agreement and precision of measurements. The clinical status of subjects (normal, glaucoma, retinal disease) was determined by "existing diagnoses" or "normal eye examinations" without detailing a specific expert panel for ground truth generation for the measurements themselves. Device operators and investigators were masked to the automated results of segmentation, but not to the clinical status.
4. Adjudication Method for the Test Set
Not applicable. This was a comparative measurement study against a predicate device, not a diagnostic accuracy study requiring an adjudication method to establish ground truth labels for the test set. Clinical status for subject inclusion was based on existing diagnoses or examinations.
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, this was not a Multi-Reader Multi-Case (MRMC) comparative effectiveness study involving human readers with and without AI assistance for improved diagnosis. This study focused on the agreement and precision of quantitative measurements between the new device and a predicate device.
6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, this study essentially assesses the standalone performance of the Canon OCT-A1 in terms of its quantitative measurements (thickness, areas, volumes) by comparing them directly to the measurements obtained from the predicate device. The segmentation (identification of retinal layers/structures) that forms the basis of these measurements is an algorithmic function of the device, and operators were "masked to the automated results of segmentation," meaning human intervention in the segmentation itself was not part of the primary comparative analysis.
7. The Type of Ground Truth Used
The "ground truth" in this context is the measurements obtained from the legally marketed predicate device (Optovue RTVue Avanti XR OCT). The study evaluates the Canon OCT-A1's agreement and precision relative to this predicate device, rather than against an independent, gold-standard ground truth like pathology or long-term outcomes data, for the quantitative measurements. The clinical categorization of subjects (normal, glaucoma, retinal disease) was based on clinical diagnosis, which provided the context for evaluating the device's performance across different patient populations.
8. The Sample Size for the Training Set
The document does not provide information on the sample size for the training set for the Canon OCT-A1's software or algorithms. The clinical study described is for validation/testing purposes to demonstrate substantial equivalence, not for algorithm training.
9. How the Ground Truth for the Training Set Was Established
The document does not provide information on how ground truth was established for any training set. It only describes the validation study for demonstrating substantial equivalence.
Ask a specific question about this device
Page 1 of 1