Table 1 Evaluation scores obtained from training Resnet-50 model from scratch and using a pre-trained model.

From: Deep learning-based classification of coronary arteries and left ventricle using multimodal data for autonomous protocol selection or adjustment in angiography

 

Resnet-50 from scratch

Pre-trained Resnet-50

Class

F1-score

AUC

F1-score

AUC

LCA

0.54

0.65

0.63

0.75

RCA

0.49

0.65

0.61

0.72

LV

0.68

0.78

0.73

0.82

Back to article page