Published September 14, 2014
| Version v1
Conference paper
Open
Extremely randomized trees based brain tumor segmentation
Authors/Creators
- 1. Medical Image Computing, DKFZ Heidelberg
- 2. Quantitative Image-based Disease Characterization, DKFZ, Heidelberg,
Description
Random Decision Forest-based approaches have previously shown promising performance in the domain of brain tumor segmenta-
tion. We extend this idea by using an ExtraTree-classi er. Several features are calculated based on normalized T1, T2, T1 with contrast agent and T2 Flair MR-images. With these features an ExtraTree-classi er is trained and used to predict di erent tissue classes on voxel level. The results are compared to other state-of-the-art approaches by participating at the BraTS 2013 challenge.