Extended Data Table 2 Performance of EchoNet-Dynamic compared with alternative deep learning architectures in assessing cardiac function

From: Video-based AI for beat-to-beat assessment of cardiac function

  1. EchoNet-Dynamic with beat-by-beat evaluation refers to the full model, which included using segmentation of the left ventricle to identify each ventricular contraction for prediction aggregation. Frame sampling refers to the performance of the base architecture on individual video clips or simple averaging across the entire video. We trained all of these architectures on the same set of Stanford videos. n = 1,277.
Back to article page