Skip to main content
. Author manuscript; available in PMC: 2021 Mar 1.
Published in final edited form as: IEEE Trans Med Imaging. 2020 Mar 5;39(9):2881–2892. doi: 10.1109/TMI.2020.2972547

Fig. 4.

Fig. 4

Segmentation results on simulated data for CGM, hippocampus and TIV across models. Top: Boxplots of different error metrics for 5 models with different augmentation: clean, classically augmented, 25%, 40%, 50% artefacts. Bottom: Bonferroni corrected pairwise Wilcoxon tests for Dice and average distance between column and row models - Green: significantly better, White: no statistically significant difference; Red: significantly worse.