(343 days)
The Neucodia system is an electrophysiological device that generates photic stimuli, and records, processes and analyzes the resultant visual evoked potential (VEP) signals for the study of central visual functions.
Visual stimuli are presented to the patient on the stimulus monitor. Visual evoked potentials are extracted from EEG epochs recorded via medical-rated EEG electrodes (not included in the device) attached to the scalp of the patient. Once the recording is complete, the digitized EEG data are processed by the software algorithm for noise filtering, artifact rejection, frequency and time domain analysis. The results are displayed on the user's monitor: EEG epochs, spectrum, Fourier components, and statistical measures (e.g., means, deviations, and signal to noise ratio).
Here's an analysis of the provided information regarding the acceptance criteria and study for the Neucodia device:
1. Table of Acceptance Criteria and Reported Device Performance
The provided 510(k) summary does not explicitly state quantitative acceptance criteria or detailed performance metrics in the way typically seen for a diagnostic or AI-driven device with specific sensitivity/specificity targets. Instead, the performance and validation section focuses on demonstrating system repeatability and reproducibility of VEP multi-variable statistics and signal-to-noise ratio.
Since no specific numerical acceptance criteria are given, the second column reflecting "Reported Device Performance" cannot be filled with quantitative values. The 510(k) focuses on demonstrating equivalence to predicate devices and the functionality of its components.
| Acceptance Criteria (Implied) | Reported Device Performance |
|---|---|
| Software Verification and Validation | Performed (stated) |
| Safety Testing (UL 60601-1, IEC 60601-2-26, EN60601-1-2, ISO15004-2) | Performed (stated) |
| Bench Testing (power supply, amplifier gain, CMRR) | Performed (stated) |
| System Repeatability (VEP multi-variable stats & SNR) | Data obtained from 10 repeated sequential tests on each subject. (Detailed results are stated to be in Section 4.2.18, User's Manual and Section 5.2.3, Device Description, which are not provided in this excerpt). |
| System Reproducibility (VEP multi-variable stats & SNR) | Data obtained from each subject having three visits for the same repeated tests using different devices, different operators, and reapplication of electrodes. (Detailed results are stated to be in Section 4.2.18, User's Manual and Section 5.2.3, Device Description, which are not provided). |
2. Sample Size Used for the Test Set and Data Provenance
The text mentions:
- "10 repeated sequential tests were performed on each subject to obtain repeatability performance data."
- "each subject had three visits for the same repeated tests using different devices and with different operators and reapplication of electrodes to obtain reproducibility performance data."
The total number of subjects (the sample size) used for these clinical tests is not specified in the provided document excerpt.
The data provenance (e.g., country of origin, retrospective/prospective) is also not specified. However, given it's a 510(k) submission to the FDA, it is likely that the study was conducted to support regulatory approval in the US, but the geographic location is not explicitly stated. The nature of the testing (repeated sequential tests, multiple visits) strongly indicates a prospective study design.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
This information is not provided in the document. The device, an electrophysiological (EEG) test system, processes signals for objective measurements (VEPs), and the "ground truth" here likely refers to the accurate capture and processing of these physiological signals rather than an expert interpretation of an image or clinical condition. Therefore, clinical expert consensus on a 'diagnosis' may not be the primary ground truth method for demonstrating system repeatability and reproducibility.
4. Adjudication Method for the Test Set
This is not applicable/not specified for the type of testing described. The study focuses on the device's ability to consistently measure electrophysiological signals, not on the interpretation of clinical outcomes where adjudication would typically be needed (e.g., for disagreements among expert readers).
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size
This is not specified and appears not to have been performed in the context of human readers improving with AI vs. without AI assistance. The Neucodia system is described as an electrophysiological measurement device, not an AI-assisted diagnostic aid for human interpretation. The study focused on the device's own measurement consistency.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
The device is an "electrophysiological device that generates photic stimuli, and records, processes and analyzes the resultant visual evoked potential (VEP) signals for the study of central visual functions." The data processing is described as being performed by a "software algorithm for noise filtering, artifact rejection, frequency and time domain analysis." The results are then "displayed on the user's monitor."
While the device's software performs processing, the description doesn't explicitly frame it as a "standalone algorithm performance" study in the context of AI-driven diagnostic accuracy (e.g., classifying a disease state). It does demonstrate the standalone processing capabilities of the algorithm in generating "VEP multi-variable statistics and signal to noise ratio" as part of its normal operation, but without an explicit "ground truth" diagnosis it aims to achieve, it's difficult to categorize it as a typical "standalone algorithm performance" in the context of a diagnostic AI. The closest analogous measure is the repeatability and reproducibility of its signal processing outputs.
7. The Type of Ground Truth Used
The "ground truth" in this context is the consistent and accurate measurement of VEP signals by the device itself. The clinical testing aims to demonstrate the device's ability to repeatedly and reproducibly produce "VEP multi-variable statistics and signal to noise ratio." The underlying biological phenomenon of Visual Evoked Potentials is the de facto ground truth that the device seeks to accurately capture and quantify.
8. The Sample Size for the Training Set
The document does not mention a training set in the context of a machine learning or AI model that requires training data. The "software algorithm" for signal processing (noise filtering, artifact rejection, frequency and time domain analysis) is presented as a deterministic algorithm rather than a learned model.
9. How the Ground Truth for the Training Set Was Established
As no training set is mentioned in the context of a machine learning model, this information is not applicable.
§ 882.1890 Evoked response photic stimulator.
(a)
Identification. An evoked response photic stimulator is a device used to generate and display a shifting pattern or to apply a brief light stimulus to a patient's eye for use in evoked response measurements or for electroencephalogram (EEG) activation.(b)
Classification. Class II (performance standards).