ECG-AI Cardiac Amyloidosis (CA) 12-Lead Algorithm (1040)

The ECG-AI Cardiac Amyloidosis 12-Lead Algorithm is software that analyzes 10-second or longer 12-lead ECG recordings using artificial intelligence to aid in earlier detection of cardiac amyloidosis in adults with symptoms or risk factors such as heart failure, nephrotic syndrome, peripheral neuropathy, arrhythmias, or aortic stenosis. It is not a stand-alone diagnostic device and must be used jointly with clinician judgment.

The device uses a machine learning-based neural network algorithm that analyzes 12-lead ECG voltage time series data and outputs a binary result (Detected, Not Detected, or Error) via API to third-party software such as EMR or ECG management systems. It is provided as a Docker container software module without its own graphical user interface and is integrated into existing clinical workflows.

The device was evaluated using software verification and validation per IEC 62304, cybersecurity testing per FDA guidance, labeling verification and validation, and human factors testing. The clinical validation used retrospective study design.

The subject device is substantially equivalent to the predicate ECG-AI LEF 12-Lead algorithm (K250652) because both are machine learning-based cardiovascular notification software with identical principles of operation, data acquisition from 12-lead ECGs, identical output formats, and identical cybersecurity compliance. Though the subject device detects a different disease (cardiac amyloidosis versus low ejection fraction), clinical performance data demonstrates it is as safe and effective as the predicate.