Table 4 Test performances of all methods on a single domain (PTB-XL) are presented. The mean and standard deviation across five random seeds are shown. Bold reflects the method with the best performance. The overall performance is the weighted average AUROC by the number of data in each site. Bold is the best and underlined is the second best

Supervised (baseline)

 Single data

0.874 ± 0.010

0.902 ± 0.013

0.890 ± 0.003

0.894 ± 0.009

 Merged data

0.903 ± 0.013

0.935 ± 0.010

0.910 ± 0.009

0.911 ± 0.008

0.914 ± 0.010

Federated

 FedAvg

0.901 ± 0.003

0.917 ± 0.007

0.909 ± 0.006

0.915 ± 0.006

0.910 ± 0.005

 FedProx

0.902 ± 0.003

0.925 ± 0.004

0.906 ± 0.008

0.906 ± 0.004

0.909 ± 0.003

Finetuning

 Small to Large

0.899 ± 0.006

0.918 ± 0.003

0.903 ± 0.005

0.905 ± 0.004

0.906 ± 0.003

 Large to Small

0.889 ± 0.005

0.926 ± 0.005

0.898 ± 0.006

0.903 ± 0.004

0.903 ± 0.004

Continual

 Small to Large

0.894 ± 0.010

0.923 ± 0.009

0.907 ± 0.010

0.904 ± 0.010

0.906 ± 0.009

 Large to Small

0.903 ± 0.006

0.929 ± 0.011

0.916 ± 0.007

0.909 ± 0.006

0.914 ± 0.007