Table 5 Approach (1) - Scenario 3: average classification evaluation results (%) over five cross-validation tests of the proposed hybrid model (CNN + ViT).

From: A hybrid segmentation and classification CAD framework for automated myocardial infarction prediction from MRI images

AI Model

ACC

SEN

SPE

F1-score

AUC

(1) ResNet50 + ViT

86.11

86.30

85.71

89.36

86.01

(2) Ensemble 1 + ViT

87.96

84.93

94.29

90.51

89.61

(3) Ensemble 2 + ViT

87.04

87.67

85.71

90.14

86.69

Back to article page