Fig. 2

ROC curve of HF in the studied subgroups. (A) Classification model – Stage A vs. Stage B groups. Confusion matrix: Accuracy 0.88, Recall 0.88, AUC ROC 0.93, F1 0.88. (B) Classification model – Stage B vs. Stage C groups. Accuracy 0.89, Recall 0.89, AUC ROC 0.97, F1 0.89. (C) Classification model – Stage C vs. Stage D groups. Confusion matrix: Accuracy 0.79, Recall 0.79, AUC ROC 0.89, F1 0.79. (D) Classification model – HFpEF vs. HFmrEF vs. HFrEF. Confusion matrix: Accuracy 0.66, Recall 0.66, AUC ROC 0.72, F1 0.64. (E) Classification model - HFpEF vs. HF with EF < 50%. Confusion matrix: Accuracy 0.88, Recall 0.88, AUC ROC 0.96, F1 0.88. (F) Classification model - HFrEF vs. HF with EF > 40%. Confusion matrix: Accuracy 0.70, Recall 0.70, AUC ROC 0.66, F1 0.68. (G) Classification model – Cluster 1 vs. Cluster 2 vs. Cluster 3 vs. Cluster 4. Confusion matrix: Accuracy 0.95, Recall 0.95, AUC ROC 0.96, F1 0.95.