Search Results
Found 1 results
510(k) Data Aggregation
(328 days)
Scanning Laser Ophthalmoscope Mirante [SLO/OCT Model]: The Mirante SLO/OCT with scanning laser ophthalmoscope and optical coherence tomography function and with Image Filing Software NAVIS-EX is a non-contact system for imaging the fundus and for axial cross sectional imaging of ocular structures. It is indicated for in vivo imaging and measurement of:
· the retina, retinal nerve fiber layer, optic disc, and
· the anterior chamber and cornea (when used with the optional anterior segment OCT adapter)
and for color, angiography, autofluorescence, and retro mode imaging of the retina as an aid in the diagnosis and management. The Image Filing Software NAVIS-EX is a software system intended for use to store, manage, process, measure, and display patient data and clinical information from computerized diagnostic instruments through networks. It is intended to work with compatible NIDEK ophthalmic devices.
Scanning Laser Ophthalmoscope Mirante [SLO Model]: The Mirante SLO with scanning laser ophthalmoscope function and with Image Filing Software NAVIS-EX is a noncontact system for imaging the fundus. It is indicated for color, angiography, auto-fluorescence, and retro mode imaging of the retina as an aid in the diagnosis and management. The Image Filing Software NAVIS-EX is a software system intended for use to store, manage, process, measure, analyze and display patient data and clinical information from computerized diagnostic instruments. It is intended to work with compatible NIDEK ophthalmic devices.
The Nidek Mirante 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 Nidek Mirante 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. The Nidek Mirante uses Fourier Domain OCT, a method that involves spectral analysis of the returned light rather than mechanic moving parts in the depth scan. Fourier Domain OCT allows scan speeds about 65 times faster than the mechanical limited Time Domain scan speeds. The Mirante utilizes Fourier spectroscopic imaging a Michelson interferometer. The interfering light of the reference light and the reflected light from the test eye obtained by the Michelson interferometer are spectrally divided by a diffraction grating and the signal is acquired by a line scan camera. The signal is inverse Fourier transformed to obtain the reflection intensity distribution in the depth direction of the patient's eve. The galvano mirror scans the imaging light in the XY direction to obtain a tomographic image. The Mirante includes scanning laser ophthalmoscope (SLO) functions as well as the OCT functions. The SLO component uses a confocal scanning system for image capture. The imaging light emitted from the laser oscillator passes through the hole mirror and enters the patient's eye. The reflected by the hole mirror and the signal is obtained by the detector. A resonant mirror and a galvanometer mirror placed in the imaging optical path scan the imaging light in the XY direction to obtain a flat surface image. The device includes Image Filing Software NAVIS-EX which is a software system intended for use to store, manage, process, measure, and display patient data and clinical information from computerized diagnostic instruments through networks. It is intended to work with compatible NIDEK ophthalmic devices.
The provided documentation describes the acceptance criteria and the study results for the Nidek Mirante Scanning Laser Ophthalmoscope and the Image Filing Software NAVIS-EX.
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are implicitly defined by demonstrating "substantial equivalence" to previously cleared predicate devices through agreement and precision analyses, and superior or equivalent image quality. The performance is reported in terms of comparisons against these predicate devices.
Nidek Mirante (OCT Component) vs. Optovue Avanti (Predicate)
| Metric | Acceptance Criteria (Implied by Substantial Equivalence) | Reported Device Performance (Nidek Mirante) |
|---|---|---|
| Agreement Analysis (Mean Difference) | Demonstrate agreement with predicate device deemed clinically acceptable. | [ILM-RPE/BM] Thickness: Higher than Avanti (10-20 µm thicker). All parameters and populations met agreement performance goals. Disc Map RNFL Thickness: Higher than Avanti (around 10 µm thicker), with the exception of TSNIT Temporal. All parameters and populations met agreement performance goals. Disc Map Optic Disc: Lower Horizontal C/D Ratio and Vertical C/D Ratio, higher Disc Area and Cup Area (All Subjects). Similar differences for Normal, lower values for Glaucoma. All parameters and populations met agreement performance goals. Cornea Radial CCT: Higher than Avanti (around 15 µm thicker). Agreement performance goals met for All Subjects, but not met for Normal and Corneal Disease populations. |
| Precision Analysis (Repeatability) | Demonstrate acceptable variation (coefficient of variation, %CV) for measurements. | [ILM-RPE/BM] Thickness: Met precision goals for all parameters and groups. Disc Map RNFL Thickness: Met most precision goals for Normal population; most met for Glaucoma except for one TSNIT Nasal and one TSNIT Temporal parameter (slightly missed). Disc Map Optic Disc: Met most precision goals for Normal and Glaucoma populations, except for Cup Area in both populations (slightly missed). Cornea Radial CCT: Met precision goals for all parameters and populations. |
| Image Quality (ACA) | Clinically useful and overall quality comparable to predicate. | Not statistically significant difference in clinical utility and overall quality compared to Avanti. |
Nidek Mirante (SLO Component) vs. OPTOS P200DTx (Predicate)
| Metric | Acceptance Criteria (Implied by Substantial Equivalence) | Reported Device Performance (Nidek Mirante) |
|---|---|---|
| Image Quality (Color Fundus) | Clinically useful and overall quality comparable to predicate. | Provided better clinical utility and overall quality compared to P200DTx (p<0.0001) for all subjects and individual populations (Normal, Glaucoma, Retinal Disease). |
| Image Quality (B-FAF) | Clinically useful and overall quality comparable to predicate. | Provided better clinical utility and overall quality compared to P200DTx (p<0.0001) for all subjects and individual populations (Normal, Glaucoma, Retinal Disease). |
| Image Quality (G-FAF) | Clinically useful and overall quality comparable to predicate. | Provided better clinical utility and overall quality compared to P200DTx (p<0.0001) for all subjects and individual populations (Normal, Glaucoma, Retinal Disease). |
General Acceptance (Safety)
| Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Safety-related issues | No safety issues related to the study devices. | One adverse event (pinguecula), determined not related to the study device. |
2. Sample Size Used for the Test Set and Data Provenance
- Test Set Sample Size: A total of 170 subjects were enrolled in the study.
- 45 subjects in the Normal group
- 46 subjects in the Glaucoma group
- 47 subjects in the Retinal Disease group
- 32 subjects in the Corneal Disease group
- 167 subjects completed the study.
- Data Provenance: Prospective, comparative clinical study conducted at one clinical site in the United States.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document mentions "Masked graders reviewing Anterior Chamber Angle (ACA) and SLO images were masked to the subject, device type, subject population, configuration order, device order and results from other graders."
However, the number of experts/graders and their qualifications are not explicitly stated in the provided text.
4. Adjudication Method for the Test Set
The document states: "Scan acceptability was by a 2-step process where the device operator identified acceptable scans and then an Investigator image reviews the scans making the final determination of which scans were acceptable or unacceptable." This implies a form of sequential review, with the "Investigator" making the final determination. It does not specify an adjudication method like 2+1 or 3+1 for resolving discrepancies between multiple graders for image quality assessment, as the number of graders is not mentioned. However, for "image quality" assessments (ACA and SLO), it states "The results from the 3 masked graders were documented," and uses "grader average." This suggests that if multiple graders were used, their averages were taken, rather than a formal adjudication process to resolve disagreements.
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
This was not a multi-reader, multi-case (MRMC) comparative effectiveness study comparing human readers with AI vs. without AI assistance. The study compared the Nidek Mirante device's performance (including image quality) to predicate devices, and involved masked graders assessing image quality, but not in the context of improving human reader performance with AI assistance.
6. If a Standalone (algorithm only without human-in-the-loop performance) was Done
The primary purpose of the study was to assess the agreement and precision of the Nidek Mirante OCT measurements in comparison with a predicate device and to assess its image quality in comparison to predicate devices for both OCT and SLO. While precision analyses evaluate the device's inherent measurement consistency, and image quality assessment involves human graders, these do not represent a standalone "algorithm only without human-in-the-loop performance" in the general sense of an AI diagnostic algorithm operating independently. The device itself is an imaging system used by humans, not an AI for diagnosis.
7. The Type of Ground Truth Used (expert consensus, pathology, outcomes data, etc.)
The study does not establish an independent ground truth (e.g., pathology, clinical outcomes) for the disease states. Instead, it uses comparative effectiveness against predicate devices and agreement/precision analysis for quantitative measurements and human-graded qualitative image quality. The study categorized subjects into Normal, Glaucoma, Retinal Disease, and Corneal Disease groups. This implies that the diagnosis of these disease states served as a basis for evaluating the device's performance within those groups, but the precise method of establishing these diagnoses (e.g., expert consensus, other gold-standard tests) is not detailed for the "ground truth" of the disease classification itself. The "ground truth" for the device's performance metrics appears to be the measurements and image quality of the predicate devices, or the consistency of the Mirante itself.
8. The Sample Size for the Training Set
The document describes a clinical study to demonstrate substantial equivalence, but it does not mention a training set sample size or the development of an AI/ML algorithm that would typically involve a separate training set. The device itself is a scanning laser ophthalmoscope and optical coherence tomography system, not inherently an AI diagnostic tool. However, the NAVIS-EX software later references a "B-scan Denoising software" which is a new function. The document mentions this function "denoises a single B-scan image to an averaged image of 120 images added," suggesting it uses a computational approach, but does not provide details of a training set for this denoising algorithm if it were an AI-based method.
9. How the Ground Truth for the Training Set Was Established
As no training set is explicitly mentioned for an AI/ML algorithm in the context of the Nidek Mirante device itself, the method for establishing ground truth for a training set is not applicable or described in the provided text. For the B-scan Denoising software, the mechanism is described as "averaging 120 images," which is an algorithmic process rather than a machine learning model requiring a ground-truth labeled training set in the typical sense.
Ask a specific question about this device
Page 1 of 1