ECG-AI Low Ejection Fraction (LEF) 12-Lead algorithm (1010)

The ECG-AI LEF 12-Lead algorithm is machine learning-based software that analyzes 10 seconds of 12-lead ECG data to detect Left Ventricular Ejection Fraction (LVEF) ≤40% in adults at risk for heart failure. It aids clinicians in earlier detection and decision-making for further cardiac evaluation, but is not a stand-alone diagnostic tool and should not be used for patient monitoring or paced rhythms.

The subject device (v2.4.0) and predicate (v2.3.0) are functionally equivalent, both using machine learning algorithms to process 12-lead ECG voltage time-series data at 500Hz digital output. Both output binary results ('Low LVEF Detected,' 'Low LVEF Not Detected,' or 'Error') to third-party software for clinician display. The subject device expands hardware compatibility to include Philips PageWriter TC50 and TC30 models while maintaining identical software principles and operational characteristics.

Not stated in this summary.

The subject device is substantially equivalent because it has identical intended use (detecting LVEF ≤40% in the same at-risk patient populations), identical diagnostic application, identical patient population, identical intended users, identical environment of use, and identical output format and hardware requirements as the predicate. The minor change in naming convention ('earlier detection' vs. 'screening') does not alter the overall intended use. Expanded ECG hardware compatibility represents only an extension of the input devices supported, not a change to the fundamental algorithm or clinical application.