Search Results

Measuring temporal characteristics of saccadic refixation responses when viewing lateral visual stimulus and identifying the individual time delays of moving the eyes toward the stimuli.

The Saccadometer system measures eye movements in horizontal axis with temporal 1kHz and spatial resolution < 5 arc min. The system incorporates a built in visual target presentation using miniature laser projectors mounted on the sensor forehead plate. Saccadometer Plus is equipped with three red laser projectors with 10° separation. Saccadometer Advanced has additional extra green laser and it allows to perform more complex saccadic task. Information about eye movement is acquired by analyzing the infrared light reflected from eye orbit. The left and right eve orbit are illuminated by two infrared LEDs, providing eve surface irradiance below 1mW/cm-(operating condition). Due to the conjugacy and synchronicity of the saccadic eye movements, the rotation of left and right eyes, can be added and averaged. The inner canthi of the left and right eyes are illuminated with the low intensity IR. The difference between the amounts of IR reflected back from the eye surfaces toward the photodetecting circuitry, carries the information about the eye position changes. The eye landmarks are taking part in generation of the eye movement signal. The corneal bulge and the limbus (the border between darker iris and white sclera). Main contributor to the eye movement signal is the corneal bulge. Being the relative eye position measurement system, it requires to define the initial eye position. It means that at the beginning of every measurement, a steady eye fixation point needs to be provided. In Saccadometer the central fixation target serves this purpose. The Saccadometer Plus and Saccadometer Advanced systems are composed of: - proprietary hardware and firmware, enclosed in - examination control unit with 2 x 1.5V AA batteries, and the - eye movement sensor with integrated miniature laser spot projectors - proprietary software application LatencyMeter. - optical fiber to USB data transmitter

The provided FDA 510(k) summary for the Saccadometer Plus and Saccadometer Advanced devices describes performance data, but does not explicitly state "acceptance criteria" in a quantitative, pass/fail manner typical for clinical trials or formal validation studies. Instead, it presents results of various tests demonstrating the device's characteristics and performance.

Therefore, the table below will list the reported device performance characteristics based on the provided text, rather than explicitly stated acceptance criteria. For questions where the information is not provided in the document, it will be stated as "Not provided."

Acceptance Criteria and Reported Device Performance

Study Details

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

   • Saccade Detection Algorithm Accuracy: Test environment allowing simulation of saccade response with given parameters on a photoelectric converter. This implies a simulated, rather than human, test set. No human subject sample size is provided for this specific test.
   • Measurement Linearity Testing: "12 subjects aged 22-69."
   • Spatial Resolution of eye movement measurement: "12 subjects (aged 22-69)."
   • Reliability study: "17 subjects aged 22-69."
   • Data Provenance: For human subject studies (linearity, spatial resolution, reliability), it's explicitly stated: "Studies were performed only outside United States." This indicates the data is retrospective or prospective, but no further detail is given. For the algorithm accuracy, it was a simulated environment.

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

   • Saccade Detection Algorithm Accuracy: The ground truth was established by the simulated "pseudosaccade" parameters in a controlled test environment. No human experts were involved in establishing ground truth for this test.
   • Measurement Linearity, Spatial Resolution, Reliability: The document does not describe the establishment of ground truth by human experts for these performance tests. The tests appear to measure the device's agreement with expected physical properties or its own consistency, rather than comparing it against a human interpretation of data.

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

   • No adjudication method is described for any of the performance tests. The tests focus on device-centric performance metrics rather than interpretative output requiring human adjudication.

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 MRMC comparative effectiveness study was done. The device (Saccadometer) is an "Eye Movement Monitor" that objectively measures characteristics; it is not described as an AI-assisted diagnostic tool that human readers would use to improve their performance in interpreting images or data.

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

   • Yes, the "Experimental verification of the saccade detection algorithm" was a standalone test of the device's algorithm performance using a simulated environment. Other tests like measurement linearity, spatial resolution, and temporal resolution also assess the device's standalone performance, though they involved human subjects whose eye movements were measured by the algorithm.

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

   • Saccade Detection Algorithm Accuracy: Simulated "pseudosaccade" parameters provided as input to the photoelectric converter.
   • Measurement Linearity, Spatial Resolution, Reliability: Ground truth for these characteristics appears to be implicitly defined by the physical or physiological processes being measured (e.g., actual target displacement for linearity, intrinsic eye movement characteristics for resolution and reliability), which the device then attempts to quantify. It's not based on expert consensus, pathology, or outcomes data.

7. The sample size for the training set:

   • The document does not mention machine learning or AI models with distinct "training sets." The device's "saccade detection algorithm" seems to be based on established algorithms for eye movement analysis rather than requiring a specifically labeled training dataset in the modern sense of deep learning.

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

   • As no training set is described for a machine learning model, this information is not applicable.

Ask a specific question about this device