FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

The Family of Crystal 20 Monitors® are designed to monitor physiological signals such as cardiovascular, neurological, and muscular for the purpose of research and diagnostic purposes. The Crystal 20 Monitor is a device that is programmable to the type of physiological signals being monitored such as EEG, EOG, EMG, ENG, ECG, PSG, airflow, pulse oximetry, respiratory effort, temperature, blood pressure, etc. The signals are communicated between the patient module and the computer unit using wireless technology based on frequencies such as but not limited to 902-928 MHz, 2.4 - 2.484 GHZ, Wireless Medical Telemetry Bands (WMTS), 608-614 MHz, 1395-1400 MHz, or 1429-1432 MHz. The communication between the patient module and computer unit can also be wired (instead of wireless). The Family of Crystal 20 Monitors® will consist of three major components:

Patient Module;
Computer Unit; and
Interface Software
The Patient Module can have up to 32 channels and the ability to either transmit only (one-way) or transmit and receive (two-way). The basic functional feature of the component is to acquire signals from commercially available electrodes / sensors that are attached to the subject, perform analog-to-digital conversion (when appropriate), encode, format, and transmit the signals to the Computer Unit. The Patient Unit will also have on-board memory capability that will permit the physiological data to be stored inside the patient unit. The Patient Module will contain no-touch connectors to enable connections to commercially available electrodes / sensors.
The Computer Unit will have the ability to only receive (one-way) or receive and transmit (two-way). The basic functional feature of this component is to receive data packets, performs error detection and correction, and then sends the data to the PC Operator interface where the data can be monitored in real time or stored and analyzed at a later time.

AI/ML Overview

This 510(k) summary is for the "Family of Crystal 20 Monitors" and, as such, focuses on substantial equivalence to predicate devices rather than providing detailed clinical study results with acceptance criteria. Medical device 510(k) submissions, particularly for devices like physiological signal monitors, often rely on non-clinical performance data and comparison to predicate devices to demonstrate safety and effectiveness.

Therefore, the requested information regarding acceptance criteria, study data, sample sizes, ground truth establishment, expert qualifications, and MRMC studies is largely not present in this document. The document describes the device, its intended use, and its classification, but does not include a performance study with acceptance criteria in the way one might find for an AI/ML-driven diagnostic device.

Here's an analysis based on the provided text, highlighting what is (and isn't) available:

Acceptance Criteria and Device Performance (Not explicitly stated for clinical performance in this document)

This document does not explicitly state specific acceptance criteria (e.g., sensitivity, specificity, accuracy thresholds) for clinical performance, nor does it report the device's performance against such criteria. The submission is a 510(k) for a physiological signal monitor, which typically focuses on demonstrating technical performance (e.g., signal quality, frequency range, safety) and substantial equivalence to existing predicate devices.

The "Predicate" section lists "Crystal Monitor Model 16, K013863 and Siesta System, K003175." This indicates the primary method of demonstrating effectiveness is likely through comparison to these legally marketed devices, asserting that the new device performs similarly and is at least as safe and effective.

2. Sample size used for the test set and the data provenance

Not applicable/Not provided in this document. This document does not describe a clinical performance study with a test set of patient data to evaluate the device's diagnostic accuracy. The submission focuses on the technical specifications and intended use of a physiological signal acquisition system.

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

Not applicable/Not provided in this document. Since no clinical performance study with a test set is described, there's no mention of experts establishing ground truth.

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

Not applicable/Not provided in this document. No clinical performance study is detailed.

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

Not applicable/Not provided in this document. The device is a physiological signal monitor, not an AI-driven diagnostic tool intended to assist human readers in interpreting images or data. Therefore, an MRMC study comparing human readers with and without AI assistance is irrelevant to this device type and is not mentioned.

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

Not applicable/Not provided in this document. This device is for acquiring physiological signals; it does not contain a standalone algorithm for diagnostic interpretation in the sense of AI/ML performance. Its function is to provide raw or processed physiological data to clinicians for their interpretation.

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

Not applicable/Not provided in this document. Given the nature of the device as a physiological signal monitor, ground truth would typically refer to the accuracy of the measurement itself rather than the accuracy of a diagnostic output. The document doesn't detail specific accuracy studies, but rather focuses on overall functionality and safety in comparison to predicate devices.

8. The sample size for the training set

Not applicable/Not provided in this document. This device is not an AI/ML system that requires a "training set" of data for algorithm development.

9. How the ground truth for the training set was established

Not applicable/Not provided in this document. As there is no training set for an AI/ML algorithm, this question is not relevant to the provided text.

Summary of what the document DOES provide:

Device Description: The Family of Crystal 20 Monitors® is designed to monitor physiological signals (e.g., EEG, EOG, EMG, ECG, PSG, airflow, pulse oximetry, respiratory effort, temperature, blood pressure) for research and diagnostic purposes.
Components: Patient Module, Computer Unit, and Interface Software.
Key Features: Up to 32 channels, one-way or two-way transmission, analog-to-digital conversion, encoding, formatting, wireless (various MHz and GHz bands, including WMTS and ISM) or wired communication, on-board memory in the Patient Module, error detection/correction in the Computer Unit.
Intended Use: Monitoring and recording of physiological signals to aid in research and/or diagnostic purposes.
Contraindication: Not intended for use as life support equipment or vital signs monitoring in intensive care units.
Predicate Devices: Crystal Monitor Model 16 (K013863) and Siesta System (K003175). The substantial equivalence claim is based on comparison to these devices.
Regulatory Information: Classified as an Electroencephalograph (21 CFR 882.1400), Class II.
K Number: K042039.

In conclusion, this 510(k) summary is typical for a device primarily focused on signal acquisition and hardware, demonstrating equivalence to already cleared devices rather than presenting novel clinical performance data through a structured study with specific acceptance criteria as would be expected for a new diagnostic algorithm or AI product.

Summary

Regulation Number and Section

§ 882.1400 Electroencephalograph.

(a)
Identification. An electroencephalograph is a device used to measure and record the electrical activity of the patient's brain obtained by placing two or more electrodes on the head.(b)
Classification. Class II (performance standards).