TABLE 7.

Ablation study results on the impact of contrastive learning.

Model	AUC	Accuracy	Precision	Recall	F1 score
KneeXNet	0.985 $\pm$ 0.003	0.968 $\pm$ 0.004	0.972 $\pm$ 0.005	0.979 $\pm$ 0.004	0.975 $\pm$ 0.004
KneeXNet-C	0.980 $\pm$ 0.004	0.964 $\pm$ 0.005	0.969 $\pm$ 0.006	0.975 $\pm$ 0.005	0.972 $\pm$ 0.005

The bold values represent the best-performing results for each respective metric or category.