Skip to main content
. Author manuscript; available in PMC: 2016 Dec 1.
Published in final edited form as: Med Image Anal. 2015 Aug 28;26(1):82–91. doi: 10.1016/j.media.2015.08.010

Figure 4.

Figure 4

Parameter optimization and sensitivity for the number of atlases fused for the initial majority vote (A), and the type of weak learner used for the AdaBoost classifiers (B). A representative segmentation using the optimized parameters can be seen in (C). Note, on (B), “*” indicates statistically significant difference, and “NS” indicates no significant difference.