Search Results

The Fetalgard 3000 Fetal Monitor is a Perinatal Monitoring System for showing graphically the relationship between maternal labor and the fetal heart rate by means of combining and coordinating display of uterine contraction and fetal heart rate measurements. Uterine contractions are monitored using an external tocotonometer or an intrauterine pressure transducer. Fetal heart rate is measured using an external pulsed Doppler ultrasound transducer or directly with a spiral scalp electrode. Maternal heart rate and respiration are measured using standard ECG electrodes. Heart rate, respiration rate, and uterine activity are presented graphically on an LCD display or chart recorder and digitally on LED displays.

AI/ML Overview

The provided text describes modifications to the Analogic Corporation Fetalgard 3000 Fetal Monitor and its substantial equivalence to a predicate device, rather than a study designed to establish new acceptance criteria or prove device performance against specific targets for a novel AI device. Therefore, much of the requested information regarding AI device evaluation is not directly applicable.

Based on the provided information, here's what can be extracted:

1. A table of acceptance criteria and the reported device performance:

This document does not specify quantitative acceptance criteria in terms of performance metrics (e.g., sensitivity, specificity, accuracy) for an AI or algorithm-driven device. Instead, the focus is on maintaining the safety and performance of a modified medical device to be equivalent to an existing predicate device.

Acceptance Criteria (Implied)	Reported Device Performance
All functions potentially affected by modifications still in compliance with original user requirements.	Demonstrated compliance with original user requirements through testing.
Safety not adversely affected.	Safety maintained through testing.
Performance not adversely affected.	Performance maintained through testing.
Performance substantially equivalent to the predicate device.	Testing of the modified device demonstrates the performance is substantially equivalent to the predicate device.

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

Sample Size for Test Set: Not specified. The document only mentions "a subset of the original verification and validation test scenarios" and the use of "both simulated input signals and tape recordings of actual physiological waveforms."
Data Provenance: The document mentions "tape recordings of actual physiological waveforms," implying real-world data, but the country of origin is not specified. It is likely US-based, given the FDA submission.
Retrospective/Prospective: The use of "tape recordings" suggests retrospective data was used, but the document does not explicitly state this.

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

This information is not provided. The testing primarily focused on verifying that device modifications did not negatively impact its functionality and that it remained equivalent to the predicate device. The concept of "ground truth" as typically applied to AI model validation, especially involving expert consensus for complex interpretations, is not detailed in this submission.

4. Adjudication method for the test set:

Not applicable/Not specified. The testing described focuses on functional verification and performance equivalence to a known predicate, not on subjective interpretation requiring adjudication among experts.

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. This submission is for modifications to a fetal monitor whose purpose is data display, not for an AI-assisted diagnostic device that would involve human readers or comparative effectiveness studies of AI vs. human performance.

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

The device described is a monitoring system that displays data for human interpretation. While it contains "system software" that "undergone several evolutionary revisions to improve the detection and display of fetal heart rate under adverse conditions," the evaluation described is not a standalone performance assessment of an AI algorithm in the absence of human involvement. The device's intended use is to "aid in assessing the well-being of the fetus," implying human interpretation.

7. The type of ground truth used:

The "ground truth" in this context appears to be the expected correct functional output of the device based on the known characteristics of "simulated input signals and tape recordings of actual physiological waveforms" and comparison against the predicate device's known behavior. It's not "expert consensus," "pathology," or "outcomes data" in the sense used for AI diagnostic tools; rather, it's about accurate measurement and display of physiological signals.

8. The sample size for the training set:

Not applicable. This submission describes modifications to an existing device and its re-validation, not the development or training of a new AI model with a distinct training set. The "system software" improvements were likely based on engineering refinements and testing rather than a formal machine learning training paradigm with a specific dataset.

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

Not applicable. As noted above, this submission does not describe the training of a new AI model where ground truth for a training set would be established. The software revisions are described as "evolutionary," suggesting iterative improvements and bug fixes based on observed performance.

Ask a Question

Ask a specific question about this device

Page 1 of 1