Scale-Space and Edge Detection Using Anisotropic Diffusion
IEEE Transactions on Pattern Analysis and Machine Intelligence
A Real-Time Algorithm for Medical Shape Recovery
ICCV '98 Proceedings of the Sixth International Conference on Computer Vision
Probability Estimates for Multi-class Classification by Pairwise Coupling
The Journal of Machine Learning Research
Atlas guided identification of brain structures by combining 3d segmentation and SVM classification
MICCAI'06 Proceedings of the 9th international conference on Medical Image Computing and Computer-Assisted Intervention - Volume Part II
Active Contour External Force Using Vector Field Convolution for Image Segmentation
IEEE Transactions on Image Processing
Hi-index | 0.00 |
In this paper a novel automatic approach to identify brain structures in magnetic resonance imaging (MRI) is presented for volumetric measurements. The method is based on the idea of active contour models and support vector machine (SVM) classifiers. The main contributions of the presented method are effective modifications on brain images for active contour model and extracting simple and beneficial features for the SVM classifier. The segmentation process starts with a new generation of active contour models, i.e., vector field convolution (VFC) on modified brain images. VFC results are brain images with the least non-brain regions which are passed on to the SVM classification. The SVM features are selected according to the structure of brain tissues, gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). SVM classifiers are trained for each brain tissue based on the set of extracted features. Although selected features are very simple, they are both sufficient and tissue separately effective. Our method validation is done using the gold standard brain MRI data set. Comparison of the results with the existing algorithms is a good indication of our approach's success.