LINQ II Insertable Cardiac Monitor, Zelda AI ECG Classification System

• K Number: K210484
• Device Name: LINQ II Insertable Cardiac Monitor, Zelda AI ECG Classification System
• Manufacturer: Medtronic, Inc.
• Date Cleared: 2021-06-11 ( 112 days)
• Product Code: MXD
• Regulation Number: 870.1025
• Type: Traditional
• Panel: CV
• Reference & Predicate Devices: K200795

Intended Use

The LINQ II ICM is an insertable automatically-activated and patient-activated monitoring system that records subcutaneous ECG and is indicated in the following cases:
• patients with clinical syndromes or situations at increased risk of cardiac arrhythmias
• patients who experience transient symptoms such as dizziness, palpitation, syncope, and chest pain that may suggest a cardiac arrhythmia

Device Description

The LINQ II Insertable Cardiac Monitor (ICM) Model LNQ22 is a programmable device that continuously monitors a patient's ECG and other physiological parameters. The device records cardiac information in response to automatically detected arrhythmias and patient-initiated activation or markings. The device is designed to automatically record the occurrence of an episode of arrhythmia in a patient. Note: Arrhythmias are classified as tachyarrhythmia, bradyarrhythmia, pause, atrial tachyarrhythmia, or atrial fibrillation. Patients may also manually record symptoms. In order to manually record symptoms, the patient will also need either the MyCareLink Heart App (patient app on mobile device) or the Patient Assistant Model PA97000. The patient can use the MyCareLink Heart App or the Patient to manually record his or her cardiac rhythm while experiencing or immediately after a symptomatic event. LINQ II ICM includes the following accessories: LINQ Tool Kit Model LNQ22TK, Reveal LINQTM Mobile Manager Model MSW002, Device Command Library Model 2692, and Instrument Command Library Model 2691. New to the LINO II ICM system is the Zelda AI ECG Classification System Models ZA400. ZA410. ZA420. included in this submission.

AI/ML Overview

The provided text describes the 510(k) summary for the Medtronic LINQ II Insertable Cardiac Monitor (ICM) when used with the new Zelda AI ECG Classification System. While a general overview of testing is provided, it does not contain the detailed acceptance criteria or the specific study results demonstrating how the device met those criteria, particularly in the format of a table of acceptance criteria and reported device performance.

The document states:

• "Performance validation testing and analysis were completed to ensure the algorithms were able to reduce false alerts from ICM detected AF and Pause episodes while retaining true alerts. All results met or exceeded the criteria in the Validation Plan."
• "The results of the testing met the design requirements and specified acceptance criteria and did not raise new safety or performance issues."

However, the specific numerical acceptance criteria (e.g., minimum sensitivity, minimum specificity, maximum false alert rates) and the corresponding achieved performance values are not listed. Also, details regarding the sample size, data provenance, expert qualifications, and ground truth establishment for the test set are largely absent.

Therefore,Based on the provided text, I cannot provide a complete answer to your request as the specific detailed information on acceptance criteria and study results (like a table of performance metrics, sample sizes for test sets, number of experts, and adjudication methods) is not included in this 510(k) summary. The document states that testing was performed and met criteria, but it does not specify what those criteria or results were.

Here's what can be inferred and what information is missing:

Information Present or Inferable:

• Device Name: LINQ II Insertable Cardiac Monitor with Zelda AI ECG Classification System
• Purpose of Zelda AI: To reduce false alerts from ICM detected AF and Pause episodes while retaining true alerts.
• Study Type (Implied): Performance validation testing to ensure the algorithms met design requirements and performance criteria.
• Ground Truth (Training): The document mentions "re-training the algorithm on data labeled following the original protocol and data labeled following an alternate protocol." This implies that the training data had labels, but the method of establishing these labels is not detailed.
• Ground Truth Type (Implied): Given the context of ECG classification for arrhythmias, the ground truth would likely be established by qualified experts (e.g., cardiologists, electrophysiologists) reviewing the ECGs.
• No Multi Reader Multi Case (MRMC) Study: The document does not mention any comparative effectiveness study involving human readers. The focus of the Zelda AI is on improving the device's automatic classification, not necessarily on human reader assistance.
• Standalone Performance: The AI system is designed to improve the LINQ II ICM's automatic detection, suggesting a focus on standalone (algorithm-only) performance in reducing false alerts.
• Predetermined Change Control Plan (PCCP): Future algorithm improvements are planned under a PCCP, with methods for re-training algorithms and defined assessment metrics, acceptance criteria, and statistical methods for performance testing of proposed changes. This suggests a robust framework for future validation, but doesn't detail the initial validation.

Missing Information (Crucial for a complete answer):

1. Table of Acceptance Criteria and Reported Device Performance: This is the most significant missing piece. The document states "All results met or exceeded the criteria in the Validation Plan," but the numerical criteria and the actual performance metrics are not given.
2. Sample Sizes Used for the Test Set: This critical detail is not provided.
3. Data Provenance for Test Set: Not specified (e.g., country of origin, retrospective/prospective).
4. Number of Experts Used for Ground Truth (Test Set): Not specified.
5. Qualifications of Experts for Ground Truth (Test Set): Not specified.
6. Adjudication Method for Test Set Ground Truth: Not specified.
7. Effect Size of Human Readers Improvement (for MRMC, if applicable): Not applicable as an MRMC study is not mentioned.
8. Sample Size for the Training Set: Not specified.
9. How Ground Truth for Training Set was Established: Broadly, it states "data labeled following the original protocol and data labeled following an alternate protocol," but specifics on expert involvement or adjudication are missing.

To answer your request fully, you would need to consult a more detailed technical report or study appended to the 510(k) submission, which is typically not publicly available in this summary format.

Regulation Number and Section

§ 870.1025 Arrhythmia detector and alarm (including ST-segment measurement and alarm).

(a)
Identification. The arrhythmia detector and alarm device monitors an electrocardiogram and is designed to produce a visible or audible signal or alarm when atrial or ventricular arrhythmia, such as premature contraction or ventricular fibrillation, occurs.(b)
Classification. Class II (special controls). The guidance document entitled “Class II Special Controls Guidance Document: Arrhythmia Detector and Alarm” will serve as the special control. See § 870.1 for the availability of this guidance document.