A Comparison of Three-Dimensional Speckle Tracking Echocardiography Parameters in Predicting Left Ventricular Remodeling

Figure 4

One of the decision trees in the resulted Random Forest model. A Random Forest model is a combination of multiple diverse decision trees. The decision tree displayed in this figure had age, 3D GLS, and BSA as classification features. The grayness in a box means the probability of a node being predicted as LVR. The parameter “sample” means the number of randomly chosen samples in this node. The parameter “Gini” measures the diversity of the samples, that is, the probability of inconsistent categories between two samples from a data set. The smaller the Gini index, the higher the purity of the sample. The parameter “class” means this node tendency of this vote.