The MIDAS Journal logo

The MIDAS Journal

Home

Automated Brain-Tissue Segmentation by Multi-Feature SVM Classification

Van Opbroek, Annegreet, Van der Lijn, Fedde, De Bruijne, Marleen
Erasmus MC
The MIDAS Journal logo

Please use this identifier to cite or link to this publication: http://hdl.handle.net/10380/3443
New: Prefer using the following doi: https://doi.org/10.54294/ojfo7q
Published in The MIDAS Journal - MICCAI 2013 Workshop: The MICCAI Grand Challenge on MR Brain Image Segmentation (MRBrainS13) .
Submitted by Annegreet Van opbroek on 2013-10-23 09:38:24.

We present a method for automated brain-tissue segmentation through voxelwise classification. Our algorithm uses manually labeled training images to train a support vector machine (SVM) classifier, which is then used for the segmentation of target images. The classification incorporates voxel intensities from a T1-weighted scan, an IR scan, and a FLAIR scan; features to encode the voxel position in the image; and Gaussian-scale-space features and Gaussian-derivative features at multiple scales to facilitate a smooth segmentation. An experiment on data from the MRBrainS13 brain-tissue-segmentation challenge showed that our algorithm produces reasonable segmentations in a reasonable amount of time.