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. 6.

Fig. 6

Per-voxel mean and uncertainty estimations on CGM (top) and hippocampus (bottom) segmentation tasks for clean (no augmentation), classically augmented and motion augmented models for a test-retest pair for which one scan is heavily artefacted. The segmentation produced by a benchmark method is shown for reference. Bottom row of each block: KL-divergence (KLD) between the probability distributions produced by each model on clean and artefacted scans.