DeepRhythmAI

DeepRhythmAI is a cloud-based software that uses artificial intelligence algorithms to analyze single- or two-lead ECG data from adult patients and detect cardiac arrhythmias. It is designed for integration with healthcare applications and medical devices to provide qualified healthcare professionals with supportive diagnostic information for review alongside clinical judgment.

The device utilizes CNN and transformer deep-learning models via a scalable API for automated ECG analysis. It detects QRS complexes, heart rate, RR intervals, supraventricular and ventricular arrhythmias, and atrial fibrillation. The new feature compared to predicate is PVC morphology grouping—the ability to cluster individual premature ventricular contractions into groups of similar morphology.

ANSI/AAMI/IEC 60601-2-47:2012/(R)2016, AAMI/ANSI/EC57:2012, and ANSI/AAMI/IEC 62304 for software development and validation. Testing included unit, integration, and system-level verification; cybersecurity assessment; and validation against a proprietary database of ECG recordings from the intended patient population.

The subject device has identical indications for use, the same automated deep-learning algorithm for ECG signal analysis, equivalent software components (API and algorithm), and equivalent classification and regulatory characteristics as the predicate (K250932). The only functional difference—PVC morphology grouping—is an incremental enhancement that does not alter the core diagnostic function or intended use.