Table 1.

From IBSR to MICCAI 2012.

Str.	FIRST	FT	No TL	TL 1	TL 2	TL 3	TL 4	TL 5	TL 6	TL 7
Tha.L	0.889 ± 0.017	0.920 ± 0.017	0.529 ± 0.251	0.905 ± 0.014	0.904 ± 0.017	0.908 ± 0.017	0.909 ± 0.015	0.910 ± 0.015	0.912 ± 0.015	0.913 ± 0.015
Tha.R	0.890 ± 0.018	0.924 ± 0.016	0.418 ± 0.220	0.903 ± 0.012	0.908 ± 0.013	0.912 ± 0.011	0.914 ± 0.011	0.914 ± 0.012	0.912 ± 0.014	0.917 ± 0.012
Cau.L	0.797 ± 0.117	0.885 ± 0.071	0.694 ± 0.090	0.861 ± 0.066	0.867 ± 0.062	0.874 ± 0.063	0.875 ± 0.063	0.878 ± 0.061	0.880 ± 0.062	0.884 ± 0.060
Cau.R	0.837 ± 0.046	0.887 ± 0.057	0.774 ± 0.050	0.870 ± 0.049	0.874 ± 0.051	0.877 ± 0.050	0.883 ± 0.053	0.881 ± 0.053	0.886 ± 0.052	0.885 ± 0.053
Put.L	0.860 ± 0.080	0.909 ± 0.023	0.884 ± 0.023	0.903 ± 0.023	0.910 ± 0.024	0.911 ± 0.025	0.913 ± 0.023	0.914 ± 0.023	0.914 ± 0.023	0.915 ± 0.024
Put.R	0.876 ± 0.060	0.908 ± 0.031	0.884 ± 0.018	0.906 ± 0.024	0.910 ± 0.023	0.912 ± 0.023	0.913 ± 0.023	0.915 ± 0.024	0.913 ± 0.025	0.915 ± 0.024
Pal.L	0.815 ± 0.060	0.873 ± 0.101	0.374 ± 0.269	0.842 ± 0.032	0.856 ± 0.028	0.861 ± 0.028	0.862 ± 0.024	0.865 ± 0.023	0.866 ± 0.024	0.866 ± 0.024
Pal.R	0.799 ± 0.088	0.874 ± 0.049	0.111 ± 0.181	0.839 ± 0.043	0.850 ± 0.041	0.853 ± 0.043	0.857 ± 0.044	0.856 ± 0.048	0.858 ± 0.050	0.862 ± 0.045
Hip.L	0.809 ± 0.014	0.871 ± 0.020	0.808 ± 0.021	0.825 ± 0.034	0.835 ± 0.033	0.846 ± 0.026	0.849 ± 0.026	0.851 ± 0.024	0.854 ± 0.027	0.856 ± 0.026
Hip.R	0.810 ± 0.022	0.869 ± 0.020	0.822 ± 0.019	0.845 ± 0.019	0.841 ± 0.027	0.854 ± 0.020	0.854 ± 0.019	0.858 ± 0.020	0.861 ± 0.020	0.863 ± 0.020
Amy.L	0.721 ± 0.054	0.832 ± 0.032	0.669 ± 0.043	0.740 ± 0.043	0.777 ± 0.032	0.800 ± 0.031	0.809 ± 0.029	0.810 ± 0.030	0.812 ± 0.029	0.812 ± 0.033
Amy.R	0.707 ± 0.052	0.812 ± 0.027	0.613 ± 0.056	0.739 ± 0.047	0.750 ± 0.046	0.766 ± 0.044	0.774 ± 0.049	0.784 ± 0.040	0.788 ± 0.038	0.789 ± 0.042
Acc.L	0.699 ± 0.081	0.790 ± 0.052	0.693 ± 0.055	0.769 ± 0.050	0.779 ± 0.050	0.784 ± 0.055	0.788 ± 0.049	0.786 ± 0.050	0.789 ± 0.041	0.792 ± 0.046
Acc.R	0.678 ± 0.089	0.783 ± 0.067	0.238 ± 0.203	0.722 ± 0.093	0.729 ± 0.091	0.759 ± 0.091	0.763 ± 0.087	0.771 ± 0.090	0.774 ± 0.084	0.765 ± 0.084
Avg.	0.799 ± 0.094	0.867 ± 0.064	0.608 ± 0.272	0.834 ± 0.077	0.842 ± 0.073	0.851 ± 0.068	0.854 ± 0.066	0.857 ± 0.065	0.859 ± 0.063	0.860 ± 0.064
wAvg.	0.706 ± 0.061	0.803 ± 0.040	0.474 ± 0.083	0.754 ± 0.047	0.766 ± 0.046	0.783 ± 0.048	0.788 ± 0.046	0.791 ± 0.047	0.794 ± 0.042	0.792 ± 0.044

Mean ± standard deviation DSC values for FIRST, full training and transfer learning with an incremental number of training images. FT (Full training) – the network is trained from scratch with the MICCAI 2012 dataset. TL X – transfer learning with X number of target volumes. No TL – tested directly on the model trained with the IBSR dataset. The structure acronyms are as follows: left thalamus (Tha.L), right thalamus (Tha.R), left caudate (Cau.L), right caudate (Cau.R), left putamen (Put.L), right putamen (Put.R), left pallidum (Pal.L), right pallidum (Pal.R), left hippocampus (Hip.L), right hippocampus (Hip.R), left amygdala (Amy.L), right amygdala (Amy.R), left accumbens (Acc.L), right accumbens (Acc.R), average value (Avg.) and weighted average DSC (wAvg.).