Search Results

The Assert-IQ™ ICM is indicated for the monitoring and diagnostic evaluation of patients who experience unexplained symptoms that may be cardiac-related such as: dizziness, palpitations, chest pain, syncope, and shortness of breath, as well as patients who are at risk for cardiac arrhythmias such as bradycardia, tachycardia, and sinus pauses. The Assert-IQ ICM is also indicated for patients who have been previously diagnosed with atrial fibrillation (AF) or who are susceptible to developing AF. The Assert-IQ ICM is intended to be inserted subcutaneously in the left pectoral region, also described as the left anterior chest wall. The Assert-IQ ICM has not been specifically tested for pediatric use.

The Assert-IQ™ ICM is designed to help physicians and clinicians monitor, diagnose and document the heart rhythm in patients who are susceptible to cardiac arrhythmias and unexplained symptoms by detecting arrhythmias and transmitting data for review. The Assert-IQ ICM system, cleared under K230286 on May 17, 2023, includes implantable and remote care components. The implantable components include the Assert-IQ™ ICM device models DM5000, DM5300, or DM5500. The remote care portion consists of the Merlin.net™ Software model MN7000 and myMerlin™ mobile apps (Android (APP1000) and iOS (APP1001)).

The subject of this premarket notification is the integration of two new artificial intelligence (AI) algorithms utilizing machine learning (ML) techniques for the evaluation of atrial fibrillation (AF) and Pause episodes within the Assert-IQ™ ICM remote care component, Merlin.net MN7000. The goal of the AI-enabled function in Merlin.net is to reduce non-actionable clinical review burden due to false Pause and false AF episodes presented for clinician review. Specifically, this premarket submission pertains to the addition of the proposed deep neural network AI models as integrated sub-components of the Merlin.net software, MN7000, resulting in MN7000 version v2.0. There are no other proposed changes to the Assert-IQ device hardware, device firmware, device detection algorithms or other components of the system cleared in K230286.

The two new AI algorithms (CARE: Classification using AI for Rhythm Evaluation) classify AF and pause EGM episodes detected by Assert-IQ ICM devices as either true or false detection. Episodes classified as "true" will be retained in the transmission data and displayed to clinicians for review in Merlin.net web application, whereas episodes classified as "false" will be removed and not displayed to the user. These two AI algorithms, CARE-AF and CARE-Pause, are designed to significantly reduce false episodes, while maintaining true arrhythmic episodes detected by the Assert-IQ devices.

Here's a summary of the acceptance criteria and study details for the Assert-IQ ICM System with AI, based on the provided FDA 510(k) clearance letter:

Acceptance Criteria and Device Performance

The core purpose of the AI algorithms (CARE-AF and CARE-Pause) is to reduce non-actionable clinical review burden due to false pause and false AF episodes while maintaining true arrhythmic episodes. The acceptance criteria are therefore focused on "relative sensitivity" and "false positive reduction."

Study Details

The document describes two primary studies for assessing the performance of the AI algorithms:

1. Retrospective Observational Cohort Study (for independent AI algorithm performance)

• Sample Size (Test Set):
  • CARE-Pause: 1498 Assert-IQ ICM patients
  • CARE-AF: 911 Assert-IQ ICM patients
• Data Provenance: Retrospective, observational cohort study. Patients were from 504 clinics across the United States (for CARE-Pause) and 360 clinics across the United States (for CARE-AF). Data was from Assert-IQ ICM patients who had AF or Pause detection over 30 days of remote monitoring post device implant.
• Number of Experts & Qualifications: Not explicitly stated. The document refers to "the overall system performance of Assert IQ with CARE-AF is assessed using data collected from the Assert-IQ post-market study (NCT06172699) comparing device detection against a Holter monitor." For the retrospective study, the ground truth establishment method implies expert review, but the number and qualifications of these experts are not provided.
• Adjudication Method: Not explicitly stated. The description mentions "AF and Pause EGM episodes detected by Assert-IQ ICM devices as either true or false detection," implying expert review to establish ground truth for these episodes. The method of achieving consensus among experts for this ground truth is not detailed (e.g., 2+1, 3+1).
• Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study: No. This study focuses on standalone AI algorithm performance relative to the predicate device's detections, and then overall system performance without explicit human-in-the-loop comparison.
• Standalone Performance: Yes. The "relative sensitivity" and "false positive reduction" metrics directly assess the independent performance of the AI algorithms (CARE-AF and CARE-Pause) in classifying episodes as true or false, relative to the existing device detection. The "overall system performance" metrics also reflect the algorithm's influence on the final output presented to clinicians.
• Type of Ground Truth: The ground truth for individual episodes was established by classifying EGM episodes as "true" or "false." This likely refers to expert consensus interpretation of the EGM data, but this is not explicitly detailed.
• Sample Size (Training Set): Not provided in the text. The document only states that "Patients whose ICM data have been utilized in algorithm training and preliminary performance evaluation were completely excluded from this study" (referring to the test set).
• Ground Truth for Training Set: Not provided in the text. It can be inferred that ground truth was established for training data in a similar manner to the test set, likely through expert review of EGM episodes.

2. Assert-IQ Prospective, Multicenter Post-Market Study (NCT06172699) - for Overall System Performance of CARE-AF

• Sample Size (Test Set): 151 patients enrolled, with 135 patients having analyzable data.
• Data Provenance: Prospective, multicenter post-market study (NCT06172699). Patients had symptomatic, drug-refractory paroxysmal or persistent AF.
• Number of Experts & Qualifications: Not explicitly stated.
• Adjudication Method: Not explicitly stated. The study compared Assert-IQ ICM AF detection against Holter assessment (up to 7 days per patient). This indicates that the Holter assessment served as a primary reference for ground truth for AF detection, likely interpreted by experts.
• Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study: No, not explicitly described as such. The study compares the Assert-IQ system with CARE-AF against Holter assessment, not human readers with and without AI.
• Standalone Performance: The "overall system performance" metrics for Assert-IQ with CARE-AF from this study represents the performance of the algorithm-enhanced system.
• Type of Ground Truth: Holter assessment (likely interpreted by experts) served as the ground truth comparator for AF detection.
• Sample Size (Training Set): Not provided.
• Ground Truth for Training Set: Not provided.

Ask a specific question about this device