DeepRhythmAI

DeepRhythmAI is a cloud-based software that automatically analyzes two-lead ECG data to detect and assess cardiac arrhythmias in adults. It is designed for integration by healthcare solution integrators into web, mobile, or other applications so qualified healthcare professionals can review and confirm automated analysis results from ambulatory ECG devices or ECG management systems. Results are advisory only and not intended as the sole diagnostic means.

The main technological difference is that the arrhythmia detection algorithm has been changed from the predicate device. Both are software-only solutions with a web API interface and an automated proprietary deep-learning algorithm for ECG analysis. Neither has a primary display or graphical user interface, and both provide the same core functions: QRS detection, heart rate determination, R-R interval detection, arrhythmia interpretation, atrial fibrillation detection, and ectopic beat detection.

Performance testing followed ANSI/AAMI/IEC 60601-2-47:2012/(R)2016 and ANSI/AAMI/EC57:2012. Software development and validation followed ANSI/AAMI/IEC 62304 and FDA's January 2002 guidance on General Principles of Software Validation. Testing also included validation against a proprietary MDG database with recordings from the intended patient population.

The device has identical intended use, same patient population (adults), equivalent monitoring environments, and functionally equivalent detection capabilities as the predicate. Although the underlying algorithm changed, it was validated against the same international consensus standards as the predicate. The performance data confirms the new algorithm meets the same safety and effectiveness requirements, and differences in technological characteristics do not raise new safety or effectiveness questions.