Abbott · Class II · Cleared Sep 17, 2025
| K-number | K251221 |
| Device name | Assert-IQ (DM5000); Assert-IQ (DM5300); Assert-IQ (DM5500); Merlin.net (MN7000) |
| Applicant | Abbott |
| Product code | MXD |
| Device class | Class II |
| Decision date | Sep 17, 2025 |
| Decision | Substantially Equivalent |
| Regulation | 870.1025 |
The Assert-IQ ICM is an implantable cardiac monitor designed to detect and monitor cardiac arrhythmias in patients with unexplained cardiac symptoms or at risk for arrhythmias. The device is inserted subcutaneously in the left pectoral region and transmits data wirelessly for clinician review. This submission adds artificial intelligence algorithms (CARE-AF and CARE-Pause) to the Merlin.net remote care software to reduce false episode alerts while maintaining detection of true arrhythmias.
The subject device maintains identical hardware, firmware, device detection algorithms, longevity, remote programming, remote monitoring, and MR conditional labeling as the predicate Assert-IQ ICM system. The only difference is the addition of two new AI algorithms (CARE-AF and CARE-Pause) integrated as sub-components of the Merlin.net MN7000 software to post-process detected episodes and filter false positives without changing the core device functionality.
Not stated in this summary.
The Assert-IQ ICM with AI is substantially equivalent because it has identical intended use, indications for use, device hardware, and operating principles as the predicate. The AI algorithms function as a post-detection filtering layer that reduces false episodes while maintaining true detections, without modifying the device's core detection mechanisms or patient safety profile. Comprehensive design verification and validation testing—including AI algorithm performance assessment on 1,498 pause episodes and 911 AF episodes, system validation with clinical workflows, and human factors engineering testing—demonstrated that the algorithms achieve high sensitivity (98.2–99.4%) and substantially reduce false positives (81–90.6%) with no safety concerns and no delay in diagnosis.
View the full FDA submission: accessdata.fda.gov