Fig. 2From: Validation of machine learning models for estimation of left ventricular ejection fraction on point-of-care ultrasound: insights on features that impact performanceLinear regression plots comparing the ML model to the reference standards. The intraclass correlation coefficient (ICC) for ML model LVEF and level III echocardiographer LVEF was 0.772 [0.501,1.000] and 0.778 [0.578,1.000] for randomized single videos by visual estimate and segmentation, respectively. The ICC for single video ML model LVEF and level III echocardiographer LVEF was 0.794 [0.173, 1.000] for visual assessment and 0.843 [0.310, 1.000] by segmentation when the expert was able to review all clips for a participant. The ICC for ML model LVEF and derived reported LVEF was 0.798 [0.143, 1.000]Back to article page