Table 5 Illustrates performance metrics, including accuracy (ACC), sensitivity (SEN), specificity (SPE), precision (PR), false positive rate (FPR), F1 scores and ROC curve (AUC), for all models in the context of the 4-chamber view.

From: Radiomics early assessment of post chemotherapy cardiotoxicity in cancer patients using 2D echocardiography imaging an interpretable machine learning study

Model

ACC

SEN

SPE

PR

FPR

F1 score

AUC

SVM

0.88

0.67

0.85

1.00

0

0.80

0.68

KNN

0.83

0.60

0.81

0.89

0.03

0.72

0.58

RUS boosted

0.71

0.53

0.76

0.60

0.19

0.56

0.50

RF

0.75

0.64

0.80

0.64

0.19

0.64

0.46

  1. *Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Random Under-Sampling Boosted (RUS boosted), Random Forest (RF).
Back to article page