Figure 2
From: Predicting Hemodynamic Shock from Thermal Images using Machine Learning
Validation of manually derived CPD for shock detection and prediction. Generalized linear mixed effects models were constructed through a standard training-testing partition. ROC curves constructed on the predictions obtained on testing set achieved 79% AUC for 0 hr detection (A) and 79% AUC for 3 hr prediction (B). ROC curves for 6 hr and 12 hr prediction are shown in Supplementary Fig. S4.
