Search Results

The ARGUS PB-3000 is a vital data acquisition unit intended to be used within or connected to a medical device or a medical system (Host System) for acquiring, analysing, and transmission of patient vitals and other pertinent clinical data of following vital data of a patient.

Indicated for the following applications:

ECG monitoring and diagnostic - The ECG monitoring function provide a conventional information about the patient's ECG rhythm, heart rate, and may be used for the diagnostic measurements, interpretation and arrhythmias detection in a medical device or medical system (Host System).

Respiration rate and apnea monitoring is indicated for pneumatic issues.

IBP - Invasive blood pressure monitoring is indicated for use in patients who require continuous monitoring of physiological pressures in order to rapidly assess changes in the patient's condition or response to therapy. It may also be used to aid in medical diagnosis.

NIBP - NIBP measurement is indicated in patients who have a risk of developing high or low blood pressure.

SpO2 - These measurements are indicated for use in patients who are at risk of developing hypoxemia, carboxyhemoglobinemia, or methemoglobinemia. This monitoring may be used during no motion conditions, and in patients who are well or poorly perfused.

CO2 - The CO2 measurement is used to detect trends in the level of expired CO2. It is used for monitoring breathing efficacy and treatment effectiveness in acute cardiopulmonary care, for example, to determine if adequate compressions are being performed during CPR or to rapidly detect whether an endotracheal tube has been placed successfully.

Respiration rate and apnea monitoring is indicated for pneumatic issues.

Cardiac Output (CO) - Cardiac Output measurement is indicated for use in patients who require a non-continuous measurement of the stroke volume and 1/min volume of the heart.

Temperature - Temperature measurement is indicated in any patient that has a risk of high or low temperature.

The ARGUS PB-3000 is a vital data acquisition unit intended to be used within or connected to a medical device or a medical system (Host System) for acquiring, and transmission of patient vitals and other pertinent clinical data. It receives vital signals from the patient through external sensors and communicates with the Host System. Data is transmitted to the Host System via network connection and without storing vital data and patient demographic data on the PB-3000. Depending on the variant of the PB-3000, the device has different modules allowing measurement of the vital parameter for ECG (monitoring and diagnostic mode, and respiration), IBP, temperature, cardiac output (CO), Sp02, CO2, and/or NIBP measurements.

The Host System is designed by a 3rd party host system manufacturer who chooses the PB-3000 variant to be implemented in their Host System. The PB-3000 communication interface allows the Host System Manufacturer to setup and use the provided functions.

The provided FDA 510(k) summary for the ARGUS PB-3000 device does not explicitly provide a table of acceptance criteria and reported device performance values in a quantifiable manner that is typically requested for AI/ML device submissions. This document is a premarket notification for a traditional medical device (a vital data acquisition unit), not a software as a medical device (SaMD) with an AI/ML component requiring a detailed performance study with ground truth and expert adjudication.

However, based on the non-clinical performance data section, we can infer some of the testing performed and the standards met, which serve as the implicit acceptance criteria for this type of device. The study described focuses on demonstrating substantial equivalence to a predicate device (ARGUS PB-1000 System, K012226) through compliance with recognized electrical safety, EMC, and performance standards for physiological monitoring equipment.

Here's an attempt to structure the available information regarding acceptance criteria and the "study" (non-clinical testing) that proves the device meets them, while acknowledging the limitations of the provided document in the context of typical AI/ML device performance reporting:

1. Table of Acceptance Criteria and Reported Device Performance

Since specific numerical acceptance criteria and "reported device performance" (e.g., sensitivity, specificity, accuracy metrics as would be seen for an AI algorithm) are not presented in this document for the overall device, the table below reflects the standards that the device met. Compliance with these standards is the "reported performance" and the fulfillment of the "acceptance criteria" in this context.

2. Sample Size Used for the Test Set and Data Provenance

• Test Set Sample Size: Not explicitly stated in terms of patient data. The testing referenced are primarily benchmarking against recognized standards using test equipment, simulators, and potentially some limited physiological setups, rather than a "test set" of patient data in the way an AI/ML model would use it. For physical device performance, compliance with standards usually involves predefined test methods and specific numbers of readings or operational cycles, but not a "sample size" of diverse patient cases as would be relevant for an algorithm's performance.
• Data Provenance: Not applicable in the context of this type of non-clinical, standards-based testing. The testing is laboratory-based and simulated, focusing on hardware and software functionality and safety according to engineering standards. There is no mention of country of origin for patient data or whether it was retrospective or prospective, as no clinical data was used for direct performance evaluation for this 510(k) submission.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

• Number of Experts: Not applicable. For standards-based testing of a vital signs acquisition unit, the "ground truth" is defined by the reference values generated by calibrated test equipment or physiological simulators as specified in the relevant international or national standards (e.g., a known heart rate from an ECG simulator, a known pressure from a pressure calibrator).
• Qualifications of Experts: Not applicable in this context. The testing would be performed by qualified engineers and technicians, not clinical experts establishing medical ground truth from patient data.

4. Adjudication Method for the Test Set

• Adjudication Method: Not applicable. The "adjudication" is compliance with engineering standards and predefined test procedures. The results are compared against the specified limits or expected behaviors detailed in these standards (e.g., "within X% deviation," "detects QRS within Y ms").

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

• MRMC Study: No. This device is a vital data acquisition unit, not an AI-assisted diagnostic tool involving human readers or interpretation of complex medical imagery/signals by AI.
• Effect Size: Not applicable.

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

• Standalone Performance: The device's performance, as evaluated against the listed standards, is inherently "standalone" in the sense that the device itself (hardware and embedded software) is tested to acquire, analyze (e.g., heart rate, respiration rate), and transmit vital signs. However, it's crucial to note its intended use: "to be used within or connected to a medical device or a medical system (Host System)." The PB-3000 acquires the data, and the Host System handles aspects like display, further analysis (e.g., arrhythmia or ST segments), and alarms. So, while its data acquisition is standalone, its full clinical function relies on integration with a Host System. No specific "algorithm only" performance metric (like for an AI model) is provided beyond general compliance with the standards for ECG, NIBP, etc.

7. The Type of Ground Truth Used

• Type of Ground Truth: Reference values from calibrated test equipment and physiological simulators as defined by the referenced national and international performance standards (e.g., electrical safety standards, specific performance standards for ECG, NIBP, SpO2 modules).

8. The Sample Size for the Training Set

• Training Set Sample Size: Not applicable. Traditional medical devices like this typically do not have "training sets" in the AI/ML sense. The embedded software and algorithms are developed using traditional engineering methods and validated against specifications and standards, not through training on large datasets.

9. How the Ground Truth for the Training Set was Established

• Ground Truth for Training Set: Not applicable. As there's no "training set" in the AI/ML context, there's no corresponding process for establishing ground truth for training. Development and validation rely on established engineering principles and adherence to standards.

Ask a specific question about this device