Table 2.

The impact of applying data augmentation on the average DICE scores.

	Without augmentation			With augmentation			Change (in %)
References	WT	TC	ET	WT	TC	ET	ΔWT	ΔTC	ΔET
Benson et al., 2018	0.82	0.64	0.59	0.82	0.61	0.56	0	−5	−5
Gholami et al., 2018	0.89	0.80	0.74	0.91	0.82	0.79	+2	+3	+7
Lachinov et al., 2018*	0.91	0.84	0.77	0.91	0.84	0.78	0	0	+1
Wang et al., 2018	0.90	0.85	0.79	0.90	0.86	0.80	0	+1	+1

For the methods reported by Lachinov et al. (2018) and Wang et al. (2018), we analyzed the best-performing models.

^*The authors verified the impact of data augmentation over the training set.