Fig. 7
From: SALT: Introducing a framework for hierarchical segmentations in medical imaging using label trees
Comparison between the ground truth (first row), SALT (second row) and Version 2 of the TotalSegmentator (third row). The comparisons across various classes and datasets are systematically organized as follows: Figures (A), (B), and (C) focus on the comparison of the brain class from the CT-ORG dataset. Figures (D), (E), and (F) examine the middle lobe of the LUNA16 dataset. In Figures (G) and (H), the analysis shifts to the pericardium from the LCTSC dataset. The bone class from the CT-ORG dataset is compared in Figures (I), (J), and (K). Figures (L), (M), and (N) represent the adrenal glands from the FLARE22 datasets. Finally, Figures (O), (P), and (Q) compare the colon class from the WORD dataset.
