Search Results

The BodyGuardian™ Remote Monitoring System detects and monitors cardiac arrhythmias in ambulatory patients, when prescribed by a physician or other qualified healthcare professional.

The BodyGuardian Remote Monitoring System is intended for use with adult and pediatric patients who are at least 29 days old in clinical and non-clinical settings to collect and transmit electrocardiogram (ECG) and other health parameters to healthcare professionals for monitoring and evaluation. Health parameters, such as heart rate and ECG data, are collected from external devices such as ECG sensors.

The BodyGuardian Remote Monitoring System does not provide any diagnosis.

The BodyGuardian Remote Monitoring System (BGRMS) is a system for recording and analyzing ECG data for cardiac arrhythmias to assist healthcare professionals, including ECG technicians at 24/7 attended analysis centers in evaluating a patient's cardiac health. Reports are generated for clinician review, that provide analysis and summary of the ECG data collected during a patient's monitoring study. Both the predicate and proposed devices, feature a modular design inclusive of outpatient cardiac telemetry (commonly called mobile cardiac telemetry (MCT)), cardiac event monitor and connected/non-connected Holter modalities. Components in the system external to the software include ECG monitors, electrodes, mobile phones and apps.

The BGRMS System includes the following main components:

• ECG monitor – a patient worn device for ECG waveform data collection and transmission, utilized with compatible electrodes
• Mobile App – applications that execute on an off-the-shelf (OTS) smartphone to communicate with the ECG monitor and the PatientCare Server for collection and transmission of data
• PatientCare – server software responsible for receiving, storing, analyzing, and displaying and reporting data gathered from the ECG monitors; includes the ECG analysis algorithm BeatLogic™
• AI-Based Device Software Functionality (AI-DSF) – Automated classification of continuous
  ECG based on the proprietary BeatLogic™ AI algorithm. BeatLogic consists of an ensemble of deep neural networks (DNNs), trained on real-world patient data and post-processing logic that combines the DNN output to produce individual beat, rhythm, and waveform classifications. This output is intended to be reviewed and confirmed by healthcare professionals to assist in diagnosis.

The provided FDA 510(k) clearance letter and summary for the BodyGuardian Remote Monitoring System (BGRMS v3.0) contains information on the device's acceptance criteria and study to prove it.

Acceptance Criteria and Reported Device Performance

The clinical validation results met all predefined acceptance criteria, though the specific criteria are not explicitly detailed in the provided document beyond "substantially equivalent performance for BeatLogic." The performance was assessed by evaluating the Sensitivity and Positive Predictive Value (PPV) for key rhythms. While specific numerical values for the acceptance criteria are not given, the reported device performance is stated as meeting these unspecified criteria.

Details of the Study Proving Device Meets Acceptance Criteria

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

• Sample Size: Not explicitly stated, but described as "real-world, randomly selected ECG records" with a demographic breakdown of 48.6% Female, 39.2% Male, 12.2% unknown gender, 50.1% < 65 years of age, 49.8% ≥ 65 years of age, and 0.1% unknown age.
• Data Provenance: "Real-world patient data" with representation across "US geographic region," indicating data from the United States. The data is retrospective as it was used to train and validate the algorithm, selected to reflect various algorithm outputs, compatible ECG device configurations, accessory types, and demographic factors.

2. 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 document states that the BeatLogic™ AI algorithm's "output is intended to be reviewed and confirmed by healthcare professionals to assist in diagnosis," but it does not specify how the ground truth for the test set was established or the number/qualifications of experts involved in this process.

3. Adjudication Method for the Test Set

This information is not provided in the document.

4. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done

• Was it done?: No, an MRMC comparative effectiveness study is not explicitly mentioned. The study focuses on the standalone performance of the AI algorithm (BeatLogic™).
• Effect size of human readers improvement: Not applicable, as an MRMC study was not described.

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

• Was it done?: Yes. The performance measurement of the BeatLogic™ algorithm involved evaluating Sensitivity and PPV, which are metrics typically used for standalone algorithm performance against a ground truth. The document explicitly states, "Performance of the algorithm was assessed by evaluating the Sensitivity and Positive Predictive value (PPV) for key rhythms across different patient subgroups." Furthermore, it mentions that the algorithm's "output is intended to be reviewed and confirmed by healthcare professionals to assist in diagnosis," implying that the performance reported is that of the algorithm prior to human review.

6. The Type of Ground Truth Used

The ground truth annotations were established based on "ground truth annotations on real-world ECG data." The method of establishing these annotations (e.g., expert consensus, pathology, outcomes data) is not explicitly stated. However, the context of cardiac arrhythmia detection strongly suggests ground truth would be established by qualified cardiologists or electrophysiologists.

7. The Sample Size for the Training Set

• Sample Size: Not explicitly stated, but described as "real-world, randomly selected ECG records" that ensured "representation across algorithm outputs, compatible ECG device configurations and accessory types and demographic factors encompassing patient age, gender, geographic location, and indication for monitoring."

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

The document states that the BeatLogic™ AI algorithm consists of "deep neural networks (DNNs), trained on real-world patient data." However, the specific method for establishing the ground truth for this training data is not explicitly provided. It is implied that this involved annotations on "real-world patient data," but the process for generating these annotations (e.g., expert review, automated processes) is not detailed.

Ask a specific question about this device